2021-10-01 01:09:55 Saijin_Naib[m]: what's the tool you're missing? From what I found in alpine's repo, there are package name xfce4-*-plugin 2021-10-01 09:32:55 hi, I am reading the installation page and it says to install `musl-locales` to get non-C locales but the package doesn't seem to exist 2021-10-01 09:33:32 have you ran apk update and everything? 2021-10-01 09:33:41 yeah 2021-10-01 09:33:49 also enabled the community repo 2021-10-01 09:34:08 "apk update" reports knowing 14928 packages 2021-10-01 09:36:25 if I search on https://pkgs.alpinelinux.org/packages it's not there either 2021-10-01 09:36:58 looks there to me? https://pkgs.alpinelinux.org/package/v3.14/community/x86_64/musl-locales 2021-10-01 09:38:51 hmm 2021-10-01 09:38:52 weird 2021-10-01 09:39:07 oh it is there 2021-10-01 09:39:18 wait it just worked 2021-10-01 09:39:19 what 2021-10-01 09:39:21 whatever 2021-10-01 09:59:44 I'm having a problem with running X11 with a radeon GPU, I've set it up with the `ati` driver as the wiki says (this is an old laptop, so it's pre-amdgpu) 2021-10-01 09:59:49 should I pastebin the x log? 2021-10-01 10:03:00 http://sprunge.us/cEcGeT 2021-10-01 10:03:05 (here it is) 2021-10-01 10:13:32 is your user in the video group 2021-10-01 10:14:48 addgroup nico video, then relogin and try again 2021-10-01 10:17:21 ah let's try that 2021-10-01 10:20:39 done that, now I'm getting another error that just says "Caught signal 6 (Aborted). Server aborting" 2021-10-01 10:20:47 but the other errors are gone 2021-10-01 10:24:25 could be a bunch of things then 2021-10-01 10:26:27 oh, on the terminal when I startx I have "Couldn't open libEGL.so.1: Error loading shared library libEGL.so.1: No such file or directory" 2021-10-01 10:29:13 but that's not showing as an error in the X log 2021-10-01 10:29:41 ah 2021-10-01 10:29:43 you want the uhh 2021-10-01 10:29:59 mesa-egl 2021-10-01 10:30:15 apk add that thing 2021-10-01 10:33:38 now my wifi has decided to just stop working 2021-10-01 10:33:47 I'll come back later, for now I'm gonna take a break from this I think 2021-10-01 10:33:56 hah 2021-10-01 10:34:01 rest well 2021-10-01 12:16:46 hi, sorry if this has come up before, but is it currently possible to do an alpine installation using glibc instead of musl libc? if i understand correctly the answer is no? (some potentially outdated wiki pages have been throwing me off a bit) 2021-10-01 12:17:09 no 2021-10-01 12:17:38 as in it's not possible, right? ok cool thx 2021-10-01 12:17:51 what's your goal ? not using glibc is one of the selling points of alpine 2021-10-01 12:17:54 heresy 2021-10-01 12:18:31 hehe. mostly just curiosity :) 2021-10-01 12:18:54 you can use `gcompat` package if you need to provide some `glibc` compatibility 2021-10-01 12:19:05 ACTION nods 2021-10-01 12:19:11 ariadne wrote a post about this 2021-10-01 12:19:18 but, please, for the love of god, don't ever mention Alpine with Glibc 2021-10-01 12:19:24 i've heard a bit about gcompat. anyway, the reason i asked was that a friend was saying it's possible, but he hadn't used alpine in a good while 2021-10-01 12:19:45 https://ariadne.space/2021/08/26/there-is-no-such-thing-as-a-glibc-based-alpine-image/ 2021-10-01 12:20:26 you can use Void with glibc 2021-10-01 12:20:35 yes 2021-10-01 12:20:37 yeah 2021-10-01 12:21:51 anything is possible, but is worth it when there are different premade solutions available? 2021-10-01 12:23:05 I mean, you can go to LFS and build anything from source 2021-10-01 12:23:45 but it's easier to install a prepackaged distro 2021-10-01 12:26:59 lots of times, people run away from a bad distro and then try to make their new distro the same as the bad one they ran from 2021-10-01 12:27:10 you could even s/distro/OS there 2021-10-01 12:30:10 fwiw i think that an alpine-like distro with glibc and other GNU userland components wouldn’t be a bad thing 2021-10-01 12:30:18 the problem is that mixing is bad 2021-10-01 12:30:43 that said glibc is basically GNU in name only anymore 2021-10-01 12:31:38 but our GNU/Linux friends deserve to have high quality package management and all of the other benefits of alpine 2021-10-01 12:31:41 :) 2021-10-01 12:32:38 but trying to mix glibc and musl ABIs will just lead to sadness 2021-10-01 12:32:50 please do yourself a favor and don’t do that :) 2021-10-01 12:40:26 thanks Ariadne :) yeah, mixing the two doesn't sound like a good time 2021-10-01 12:43:34 you can wind up in states where some parts are provided by glibc and other parts are provided by musl 2021-10-01 12:43:41 which is, umm, not ideal :D 2021-10-01 12:45:14 ACTION nods 2021-10-01 12:51:02 hello folks, how can I retrieve the checksum of individual APK packages? 2021-10-01 12:51:48 the apk manpage does not seem to show a way 2021-10-01 12:54:52 apk fetch $package -s | sha512sum 2021-10-01 12:54:54 like this? 2021-10-01 12:56:37 does fetch download or grab it from the cache? 2021-10-01 12:57:15 i think it downloads, not sure 2021-10-01 12:57:23 depends i think 2021-10-01 12:57:32 Misthios: right 2021-10-01 12:57:53 are we talking about a whole package or a file in the package 2021-10-01 12:58:07 Misthios: the package itself 2021-10-01 12:58:28 I need to provide provenance on the packages installed 2021-10-01 12:58:48 so I think I'll need to check if the files from the package are unmodified 2021-10-01 12:59:19 apk verify? 2021-10-01 13:00:17 that requires me to list the content of the packages 2021-10-01 13:00:57 it would be good to be able to provide it a package name 2021-10-01 13:01:36 apk verify path to apk does show all the content? 2021-10-01 13:01:39 or am i missing something 2021-10-01 13:02:21 ah 2021-10-01 13:05:43 thanks Misthios and vyivel, I' 2021-10-01 13:05:50 I'll see what I can hack up 2021-10-01 13:05:55 gl 2021-10-01 13:34:14 small followup question, is there a python package to interact with apk? 2021-10-01 13:35:44 not as far as i know? but u could use python ro run shell commands to interact with apk 2021-10-01 13:36:09 alright, thanks! 2021-10-01 14:16:20 does busybox have a tool to see how much bandwidth is used? (similar to what nethogs does) 2021-10-01 14:20:53 watch grep -v '^0$' /sys/class/net/*/statistics/* 2021-10-01 14:21:00 or just install nethogs 2021-10-01 14:22:18 Hello71: oh cool :) 2021-10-01 14:51:22 lieter: There is a Lua module for interacting with apk: https://pkgs.alpinelinux.org/package/edge/main/x86_64/lua5.3-apk 2021-10-01 14:55:16 how many MB pr sec download is normal these days? 2021-10-01 14:55:36 That's a very contextual question 2021-10-01 14:56:10 seems im getting 2.6MB/s 2021-10-01 14:56:31 (kinda free throught the apartment internet) 2021-10-01 14:57:00 ~25 mbit 2021-10-01 14:57:43 six quarts 2021-10-01 15:49:18 "Saijin_Naib: what's the tool you..." <- That's a great question. Those plugins are almost entirely for the panel, not the Control/Settings area. I seem to be missing Color/Monitor management, Network connection management, Bluetooth Management at the least. I'm comparing to XFCE as-shipped on other distros, no idea what they're including that integrates nicely with the XFCE Control panel (which is why I'm digging). 2021-10-01 15:49:18 gnome-control-center provides all of this inbox, which was pretty nice 2021-10-01 15:49:31 Good morning, I installed podman, created a container, and when I run this message appears. 2021-10-01 15:49:36 notwolf:~$ podman run c3ef3b9958a1 2021-10-01 15:49:54 systemctl: not found 2021-10-01 15:50:20 https://wiki.archlinux.org/title/Podman 2021-10-01 15:50:30 Are you running as root? 2021-10-01 15:50:40 Is it possible to fix it? 2021-10-01 16:01:08 well, systemctl is not going to work on Alpine Linux 2021-10-01 16:07:41 why would podman call systemctl on alpine when it isn't built with systemd support (because it cant) 2021-10-01 16:12:06 is it a bad idea to just format sdb if you want one partition instead of making sdb1 and formating it? 2021-10-01 16:18:05 not really 2021-10-01 16:20:37 "Good morning, I installed podman..." <- How did you install `podman`? 2021-10-01 16:21:05 I'm pretty sure that `podman` installed via `apk` doesn't call `systemctl` 2021-10-01 16:28:30 cool 2021-10-01 16:29:37 PJ[m]: wiki alpinelinux 2021-10-01 16:33:33 https://wiki.alpinelinux.org/wiki/Podman 2021-10-01 16:33:51 added the rc-update add podman, and /etc/init.d/podman start 2021-10-01 16:35:30 then I'm guessing the image you are trying to use tries to do something with `systemctl`? 2021-10-01 16:36:05 do you get same error when you try to run `podman run --rm ubuntu:latest`? 2021-10-01 16:38:15 or `podman run --rm hello-world` like wiki suggests 2021-10-01 16:39:35 (podman run --rm ubuntu:latest) No errors 2021-10-01 16:40:00 (hello-word)Error: Error initializing source docker://hello-word:latest: Error reading manifest latest in docker.io/library/hello-word: errors: 2021-10-01 16:40:00 denied: requested access to the resource is denied 2021-10-01 16:40:00 unauthorized: authentication required 2021-10-01 16:40:27 huh 2021-10-01 16:40:33 access denied 2021-10-01 16:41:37 not sure why it gives access denied, but nonetheless that confirms the image you tried in the beginning depends on `systemctl` for *some reason* 2021-10-01 16:44:26 (hello-word) root ok 2021-10-01 17:08:34 It was giving an error because it's already running, but it doesn't show up in the podman ps. 2021-10-01 17:08:40 https://pastebin.com/WuwhX4Rn 2021-10-01 17:15:42 hmm grabbing the 'df' used value from a hdd im copying almost 1T to, waiting 1m and grabbing it again i get 2476, but 'df -h' show ca 5G differens .. how is 2476 made into 5G? 2021-10-01 17:16:09 reserved space 2021-10-01 17:17:27 nothing is easy ;| 2021-10-01 17:29:04 wolf_: "it doesn't show up in the podman ps" Note that root podman and rootless podman are separate. `podman ps` and `sudo podman ps` will give different results. 2021-10-01 17:31:27 Yes. I don't have sudo. I'm using the wolf user. At ps appears. 2021-10-01 17:42:37 Do you have the cgroups v2 fs mounted? There's an rc service for it called cgroups. 2021-10-01 17:56:53 xordspar0: Yes 2021-10-01 18:49:29 Got it, is working, and now how to access? Hahaha 2021-10-01 18:49:36 https://pastebin.com/p2YrMtmR 2021-10-01 18:50:10 No 127.0.0.1:8069 does not access 2021-10-01 19:03:06 wolf_: How did you start it? You need to do `podman run -p 8069:8069 ...` to expose the port outside the pod. If you don't want to expose it outside, you could also try `podman exec -it mycountainer sh` to get a shell inside the container and test from there. 2021-10-01 19:03:25 -p binds to 0.0.0.0 on the host, so be careful. 2021-10-01 19:18:29 the diff of df has slowed down to 0-500 pr min (from 2500) ... 2021-10-01 19:24:00 hi, I'm having some problems connecting to wifi on a pretty much fresh alpine install, it used to work (earlier today) but now if I try to start wpa_supplicant I get the error "could not set interfact wlan0 flags (UP): No such file or directory" 2021-10-01 19:24:10 s/interfact/interface/ 2021-10-01 19:33:17 wlan0: failed to initialize driver interface 2021-10-01 19:36:43 when booting it says ip: ioctl 0x8914 failed: No such file or directory 2021-10-01 19:37:36 the wifi card has the PCI ID 14e4:1692 2021-10-01 22:01:38 "the wifi card has the PCI ID 14e..." <- Are you positive? 2021-10-01 22:01:38 https://pci-ids.ucw.cz/read/PC/14e4/1692 2021-10-02 02:22:39 Is there any way for apk to install a package from a different architecture (eg. x86_64 from arm64)? 2021-10-02 02:23:33 On a similar note, can the dynamic libraries from another architecture be installed? (I want to use the packages using QEMU user mode emulation) 2021-10-02 02:26:20 there's an --arch flag, and --root to manage a separate root 2021-10-02 02:26:56 i don't know if it actually works though (: 2021-10-02 02:34:14 Thanks, I'll try later. But can the dynamic libraries be installed? 2021-10-02 02:35:36 use a chroot 2021-10-02 02:37:52 they're packages like any other, just manage the --root correctly 2021-10-02 02:45:11 Hello71: Is it possible to add binfmt_misc entries to just have a few different arch packages instead of a whole chroot? I guess I could symlink the packages from the --root to /usr/bin? 2021-10-02 02:45:55 if you need to ask, you shouldn't do it 2021-10-02 02:46:38 how does your plan deal with identical file names (the vast majority of them) 2021-10-02 02:48:44 I was planning to install packages that don't exist on arm64 such as nextcloud so I don't think identical file names should be a problem. 2021-10-02 02:50:41 nextcloud is noarch 2021-10-02 02:51:08 and imo web applications should be installed outside of package manager, but that's a separate issue 2021-10-02 02:51:46 ignoring those, how will you handle the package's dependencies 2021-10-02 02:52:41 why do i get the feeling the actual answer is "enable community repository" 2021-10-02 02:52:47 nextcloud exists on arm64 2021-10-02 02:53:06 Oops, sorry for some reason I didn't see the nextcloud packages on arm64 earlier. 2021-10-02 02:53:52 Is all the nextcloud functionality available on arm64? 2021-10-02 02:54:04 such as 2021-10-02 02:54:51 you could just install it and find out 2021-10-02 02:55:35 psykose: Ok, thanks 2021-10-02 09:00:54 hi alpine 2021-10-02 09:02:14 i just installed now alpine, id like to use without GUI, so id like install file browser 2021-10-02 09:03:01 ls/cd :P 2021-10-02 09:03:04 a gui file browser, without grapghical interface like gnome or kde 2021-10-02 09:03:17 try ranger or nnn 2021-10-02 09:03:25 or vifm or midnight commander 2021-10-02 09:03:39 sudo apk add ranger ? 2021-10-02 09:03:51 i guess so 2021-10-02 09:03:59 ok great 2021-10-02 09:04:03 ill try 2021-10-02 09:04:24 i seen i have to add repository as well 2021-10-02 09:04:34 Has anyone had any luck getting Plymouth working on Alpine? 2021-10-02 09:04:42 but wiki isnt clear to me 2021-10-02 09:04:53 The community repository 2021-10-02 09:04:58 in /etc/apk/repositories 2021-10-02 09:05:15 can u post exactly what have i to type? 2021-10-02 09:05:25 You need to edit that file with an editor 2021-10-02 09:05:31 nano 2021-10-02 09:05:38 sudo apk add nano 2021-10-02 09:05:42 i know nano 2021-10-02 09:05:48 apk add nano 2021-10-02 09:06:04 then sudo nano /etc/apk/repositories 2021-10-02 09:06:14 but i dont know what i have to type on etc/apk/repository 2021-10-02 09:06:29 Did you open it? 2021-10-02 09:06:47 i cant right now, because im using keyboard here 2021-10-02 09:06:58 so it switched off 2021-10-02 09:07:06 switch off? 2021-10-02 09:07:13 pc 2021-10-02 09:07:23 oh ok 2021-10-02 09:07:25 pc poweroff 2021-10-02 09:07:31 the line should already be there, but commented out 2021-10-02 09:07:47 Remove the # before the line with community, and save it 2021-10-02 09:07:51 i try to switch it on 2021-10-02 09:13:34 i have 2 lines with community 2021-10-02 09:13:49 v3.14 2021-10-02 09:13:53 Do what ikke said on the line without 'edge' 2021-10-02 09:13:56 and edge 2021-10-02 09:14:24 done 2021-10-02 09:14:44 Then apk update 2021-10-02 09:16:40 i installed ranger 2021-10-02 09:17:00 how close it? i tryed ctrl + q 2021-10-02 09:17:07 but nothing 2021-10-02 09:17:40 just q 2021-10-02 09:17:44 try ':quit' 2021-10-02 09:18:47 just q 2021-10-02 09:19:27 installing firefox 2021-10-02 09:19:39 cross fingers 2021-10-02 09:20:42 Error: no Display enviroment variable specified 2021-10-02 09:20:58 You do not have a graphical environment? 2021-10-02 09:21:06 no 2021-10-02 09:21:19 then how do you expect to run firefox? 2021-10-02 09:21:50 how add graph enviroment? kde maybe 2021-10-02 09:22:23 I don't have experience with KDE myself 2021-10-02 09:22:45 i suppose it's in plasma-desktop package 2021-10-02 09:23:02 thank for answers anyway 2021-10-02 09:23:10 though if your original intention was to use tty you might want to install something more lightweight 2021-10-02 09:23:24 im not very polite for my translation 2021-10-02 09:23:46 M1e[m]: no worry 2021-10-02 09:23:48 uhM, just testing ranger and can't quit now: Not quitting: Tasks in progress: Use `quit!` to force quit 2021-10-02 09:23:53 how can I see what it's doing? 2021-10-02 09:24:24 i think i need firefox. so i need graph enviroment 2021-10-02 09:25:05 donoban: not sure 2021-10-02 09:25:18 M1e[m]: I started to use qutebrowser 2021-10-02 09:26:59 w Opens the task window where you can view and modify 2021-10-02 09:27:13 it seems stuck trying to get the preview of some json files 2021-10-02 09:28:35 apk add gnome 2021-10-02 09:33:08 startX 2021-10-02 09:33:17 not found 2021-10-02 09:33:32 gnome-session? 2021-10-02 09:33:36 and "startx"? 2021-10-02 09:33:36 or something like that 2021-10-02 09:34:07 setup-xorg-base 2021-10-02 09:34:16 oh yes 2021-10-02 09:36:28 not working 2021-10-02 09:37:07 i tryed all commands 2021-10-02 09:38:04 https://wiki.alpinelinux.org/wiki/Gnome_Setup 2021-10-02 09:38:51 check the Prerequisites 2021-10-02 10:28:35 Hey, is there a lighter browser than firefox that supports js and is usable by common people ? 2021-10-02 10:29:05 konqueror? 2021-10-02 10:29:05 (which is intended to run with lxqt) 2021-10-02 10:30:46 trying that 2021-10-02 10:30:58 (on a raspberry pi 3 I think) 2021-10-02 10:31:18 probably needs a shitload of kde packages 2021-10-02 10:32:13 space is not really a problem, alpine is installed on a 32gb card, the issue is the cpu and the ram 2021-10-02 10:33:08 maybe qutebrowser, though i've not used it 2021-10-02 10:33:58 that's what I use for myself, I wouldn't recommend it to ppl who don't know vi-style uis :D 2021-10-02 10:35:45 wasn't sure if you could use it with a mouse, too 2021-10-02 10:36:08 you can, but people will be confused 2021-10-02 11:53:50 btw, is it possible to totally get rid of pulse on alpine ? I've read somewhere that all packages were compiled for alsa. The rpi will be used only to play soundcloud/youtube music + possibly as a bt speaker. 2021-10-02 11:57:30 i hacent gotten bluetooth working with only bluez-alsa 2021-10-02 11:58:02 normal audio works fine though, without pulse 2021-10-02 12:04:52 ALSA can work without pulseaudio 2021-10-02 12:05:56 yea 2021-10-02 12:06:44 probably not true the other way around, since pulseaudio is just a wrapper around alsa, iirc 2021-10-02 12:25:49 yup, but pulse has some annoying stuff, I'm using sndio personally 2021-10-02 13:08:47 mildly interesting https://pr0gramm.com/static/4778619 2021-10-02 13:09:25 Epiphany isn't awful for light, but it is finicky 2021-10-02 13:10:09 You'll need GDM running and eudev/udev for gnome to behave. Might need to add your user to audio, video, input groups as well 2021-10-02 13:10:47 But GDM has been broken for me for over a month now, so I'm lightdm and Xfce4 now 2021-10-02 15:55:30 When one is switching from wpa_supplicant to iwd, does one take the interfaces to be managed out of /etc/interfaces like they do when going to NetworkManager? 2021-10-02 15:55:49 (or GNOME Control Center, I'm not sure if Network Manager itself actually requires that) 2021-10-02 15:57:32 depends on if you set enablenetworkconfiguration=true 2021-10-02 15:59:30 Ah, okay. I'm trying to figure out what all needs to be configured. I'm not sure if I'm just not following the Alpine Wiki guide properly or not, but last time I tried IWD it would drop connection every 5 seconds 2021-10-02 15:59:42 I thought it was some conflict with DHCP/DNS 2021-10-02 15:59:52 Since it couldn't resolve any hostnames for ping 2021-10-02 16:03:21 Does that then mean that one stops the networking rc-service? 2021-10-02 16:03:33 (along with removing/stopping wpa_supplicant) 2021-10-02 16:05:36 xordspar0: Thanks, but it doesn't connect to the bank without "pod", and with "pod" it doesn't release the port, you still have to study a lot. 2021-10-02 16:09:50 Saijin_Naib: yes, remove the interface from `/etc/network/interfaces` like you do normally, and remove the networking service. NetworkManager takes it over. Make sure you start the iwd service as well and configure NetworkManager to use iwd 2021-10-02 16:17:36 can fbset be used to expand desktop in xorg? 'fbset -yres 1000' seem to be accepted but does not reduce my screen (which is 1080) 2021-10-02 16:26:14 no 2021-10-02 16:27:21 if you want 80px of black bars there are probably some weird xrandr commands 2021-10-02 16:28:07 probably the same ones one uses to fix overscan/underscan, you just force an underscan to get some reduced screen 2021-10-02 16:35:12 yes, i was curious if fbset could replace xrandr for this 2021-10-02 16:48:25 not an expert, but once you are in a graphical session there is no fb anymore, it's x writing directly to the screen (hence drm), or at least that's my guess 2021-10-02 16:52:42 yeah, i dunno, just read the busybox man page, and looked like it _might_ have this function :] 2021-10-02 20:06:09 Hi everyone. I'm trying to configure podman on Alpine. When I try to run any container I get error: Error: OCI runtime error: invalid file system type on '/sys/fs/cgroup' 2021-10-02 20:07:00 I can't find any documentation about how to configure cgroups on Alpine edge. 2021-10-02 20:07:16 This page describes how to setup libcgroup on 3.8 https://wiki.alpinelinux.org/wiki/Docker 2021-10-02 20:07:30 but it seems that libcgroup is deprecated. Please assist 2021-10-02 20:13:30 I have cgroupfs mounted via fstab 2021-10-02 21:13:45 ngortheone: there is a cgroups service that sets up cgroups, and in /etc/rc.conf you can set rc_cgroup_mode for unified/legacy/hybrid for v2/v1/both 2021-10-02 21:15:24 there's a chance your fstab entry was wrong, but there's no real point to do it yourself since you can just add that service 2021-10-03 00:11:19 <_aelm> Hi y'all! I'm trying to get alpine booting on an rPi 0W. The install script is running just fine, but booting after is giving a kernel panic. Linux is saying that it's unable to mount the root fs on unknown-block(0 0). Are there any simple things I should be looking for? 2021-10-03 00:18:33 that sounds like you're missing a module for the type of filesystem at the root, or the module for storage drivers 2021-10-03 00:20:14 as in it should be in modules= on the cmdline 2021-10-03 00:21:00 <_aelm> psykose: the filesystem (ext4) is there. What would I need for the storage drivers to boot from the sd card? 2021-10-03 00:21:32 sd-mod i think 2021-10-03 00:21:39 <_aelm> also there by default, interesting 2021-10-03 00:22:10 there is a chance they're missing from the initramfs but i doubt that, so it's probably something else then 2021-10-03 00:23:17 <_aelm> yeah i wouldn't expect them to be off of the initramfs either, i'll keep searching 2021-10-03 00:25:05 unknown-block(0,0) only occurs if initramfs is totally missing 2021-10-03 00:26:07 sd-mod is for SCSI disks, not SD cards 2021-10-03 00:27:53 <_aelm> Hello71: ah, thanks! initramfs was certainly generated in the install script, so it would be very odd for it to be missing. If the sd card driver isn't sd-mod, it's probably not being loaded. The only other module I have is usb-storage. Do you know what I should have for the sd card? 2021-10-03 00:28:23 modules are loaded from initramfs 2021-10-03 00:28:34 your initramfs *file* may be present but not the command line option 2021-10-03 00:30:58 <_aelm> Hello71: well that could be an issue. I'm really not familiar with the boot process here at all, what should I be doing to fix that? 2021-10-03 00:31:48 what exactly did you do to "install script" 2021-10-03 00:32:35 <_aelm> I just ran setup-alpine as the wiki suggested, and installed in sys mode on the sd card. 2021-10-03 03:26:48 PureTryOut (matrix.org), Thanks so much! That was the missing piece! networking needed to be removed from rc-update. iwd working great now with NetworkManager. I'm doing migrations as I learn of them (sudo -> doas, wpa_supplicant -> iwd, and hopefully pulseaudio -> pipewire) 2021-10-03 03:59:06 Does anyone know if Alpine will be using the in-kernel ksmbd/cifsd and/or in-kernel Paragon NTFS read/write (if they aren't already)? I think I'd like to use those for what I understand should be pretty significant gains in efficiency 2021-10-03 04:11:21 don't remember if paragon made it in, but if it does i don't see why it wouldn't be enabled as a module 2021-10-03 04:12:00 ntfs is already enabled, so 2021-10-03 04:14:06 This may be dumb, but if we have the module, do we still need the utils/tools packages? Yeah, right? 2021-10-03 04:14:24 I just wouldn't get the fuse-ntfs-3g package since it would be in-kernel isntead of userspace right? 2021-10-03 04:14:31 yes 2021-10-03 04:14:42 Ah, awesome. 2021-10-03 04:16:02 Does anyone run pipewire instead of PulseAudio? I'm seeing it can reduce CPU load on audio playback which sounds quite welcome since rhythmbox/lollypop/parole/et al sit at about 7% CPU for mp3 decode from local ext4 volume, which seems high to me 2021-10-03 04:17:01 not sure which utils/tools you mean, you can mount say ext4 without e2fsprogs, etc (pretty sure), they're usually for fsck/fs specific utilities, but optional 2021-10-03 04:17:05 i use pipewire and it works fine 2021-10-03 04:19:37 I may be confusing things... I thought I needed the whateverfsname-utils/progs to interact with that FS, but I could be very wrong. 2021-10-03 04:19:46 What DE do you run with PipeWire, if I may ask 2021-10-03 04:19:55 none 2021-10-03 04:20:10 you need those to create, resize, or change some other metadata about the FS, but you don't need them to mount or use them 2021-10-03 04:20:12 just sway 2021-10-03 04:53:46 Oh, very minimal. Nice! 2021-10-03 04:54:04 Sorry for the delay. Alpine trapped and then IWD wasn't playing nice anymore 2021-10-03 04:54:22 Now I'm back to using networking to manage my interfaces but IWD for connectivity, and it seems to be working 2021-10-03 04:54:44 Better than falling back fully to wpa_supplicant, but networkmanager is not friendly in my setuo 2021-10-03 05:01:03 Gotta do some latex. Texlive? 2021-10-03 05:02:48 seems to exist, what about it 2021-10-03 05:17:25 psykose: Is that what you use? 2021-10-03 05:17:49 ah, personally i've never used latex 2021-10-03 05:17:57 didn't read that as asking for suggestions, sorry 2021-10-03 06:05:05 no verries 2021-10-03 06:40:48 anyone using telegram-desktop? is it possible to change font size? (pretty sure ive done it before ,but cant find the option now) 2021-10-03 06:42:22 wasn't it qt? QT_FONT_DPI or QT_SCALE_FACTOR (whole ui) should change it 2021-10-03 06:49:13 i think it was an option available in the app itself, but maybe im misremembering 2021-10-03 06:49:27 not sure how to change those vars 2021-10-03 06:49:53 ahh, then no idea 2021-10-03 06:50:15 for those it's just environment, i.e. QT_FONT_DPI=96 telegram-xyz 2021-10-03 06:53:32 psykose: oh this worked, and i guess its simpler :] 2021-10-03 07:01:12 Saijin_Naib: I use pipewire as a daily driver and it works great. I don't know anything about CPU usage of it though 2021-10-03 08:02:49 PureTryOut (matrix.org): you mean on Alpine? 2021-10-03 08:03:33 Yes ofc 2021-10-03 08:06:02 that's very cool! I've always thought of Alpine like a super minimal system not really suited for desktops but now I might try to create a desktop system with Alpine 2021-10-03 08:23:50 is that something that actually has successfully been done before btw? 2021-10-03 08:24:29 Sure 2021-10-03 08:57:38 ominous music starts playing in the background 2021-10-03 09:33:36 alpine is far from super minimal 2021-10-03 09:33:45 getting a usable desktop on it is 2021-10-03 09:33:47 pretty easy 2021-10-03 10:34:11 Just gotta know what u need and ur fine 2021-10-03 15:20:42 "alpine is far from super minimal" <- so what would you consider super minimal? 2021-10-03 15:27:33 kiss 2021-10-03 15:30:45 hmm, not even Gentoo? 2021-10-03 15:35:44 gentoo is far from minimal 2021-10-03 15:35:46 lol 2021-10-03 15:35:53 portage is massive 2021-10-03 15:36:09 and way too complicated 2021-10-03 15:46:47 (and python) 2021-10-03 18:38:54 Here is minimal: https://100r.co/site/uxn.html 2021-10-03 18:41:39 https://wiki.xxiivv.com/site/varvara.html 64k RAM, 256 Bytes Zero Page, 256 Bytes Stack and Return Stack, 4 Colors... 2021-10-03 18:41:45 thats not a linux distro 2021-10-03 18:41:55 devine uses manjaro ;) 2021-10-03 18:42:06 @eris linux can't be minimal because of the kernel. 2021-10-03 18:42:21 oh i know 2021-10-03 18:42:28 it's called relativity 2021-10-03 18:42:43 Hey,I'm trying to setup a raspberry pi 3 as a bluetooth speaker. Technically, it works, however, the sound is very far from great (I hear only noise). Any tips ? 2021-10-03 18:42:58 I'm using pulse for that 2021-10-03 18:43:10 things can have qualities relative to each other, even if said quality is overall not present 2021-10-03 18:43:21 xaltsc: what guide did you use to install (if any)? was it recent? 2021-10-03 18:43:35 :p 2021-10-03 18:43:47 I am currently on Manjaro too... definitly not minimal. 2021-10-03 18:43:49 frag: Multiple guides, amonst them that on the wiki, but it's aimed at the RPi 4 2021-10-03 18:44:10 so it was quite hard? 2021-10-03 18:44:24 frag: Took me several hours 2021-10-03 18:44:34 last i checked i think it had a warning of some problem, the wiki that is 2021-10-03 18:44:48 okok 2021-10-03 18:45:03 xaltsc: check if asound.conf helps. 2021-10-03 18:46:23 streampunk: I'm not using alsa at all 2021-10-03 18:46:26 xaltsc: if if type "asound.conf" in google it autocompletes to "asound.conf raspberry pi" ... :-) 2021-10-03 18:46:33 (besides from pulseà) 2021-10-03 18:46:54 xaltsc: you cannot not use alsa. alsa is always there. 2021-10-03 18:47:24 xaltsc: ok, not completely true, but I guess you don't have a firewire soundcard. 2021-10-03 18:47:29 Maybe (one of) the problem(s) is that I no external devices ? 2021-10-03 18:48:01 streampunk: Yup, just the RPi, and the sound goes through HDMI 2021-10-03 18:48:50 pulse can't use any soundcard directly. jackd can use firewire soundcards directly. 2021-10-03 18:50:36 the typical problem is that the soundcard only supports one bitrate and alsa is not using that bitrate. fix the rate with a dmix in asound.conf. 2021-10-03 18:52:14 https://www.alsa-project.org/wiki/Asoundrc#dmix 2021-10-03 18:55:39 But I dont have any pbs playing music from youtube or soundcloud though 2021-10-03 19:06:18 streampunk: any idea what parameters I should set ? 2021-10-03 19:10:14 xaltsc: https://git.io/JatYW try 48000 and 44100 for $RATE and the name of the card is found with "aplay -L" 2021-10-03 19:10:58 streampunk: do you have an explanation as to why youtube sound plays fine though ? 2021-10-03 19:12:13 xaltsc: oh, sorry I missed "bluetooth", but I idea for that too. 2021-10-03 19:15:10 xaltsc: for blutooth you have to change the codecs list. I use pipewire and there is set: `bluez5.codecs = [ sbc aac ldac aptx aptx_hd ]` 2021-10-03 19:17:15 xaltsc: I found this: https://wiki.debian.org/BluetoothUser/a2dp#AptX.2C_LDAC.2C_and_AAC_codecs_are_not_available_with_PulseAudio but I don't know if that is still true. 2021-10-03 19:19:31 xaltsc: here is the project that added the high-quality codecs to pulseaudio: https://github.com/EHfive/pulseaudio-modules-bt#configure seems upstreamed, but maybe the module config is still valid. 2021-10-03 19:21:15 streampunk: just to be clear: audio works, it's just bad quality and stuttering 2021-10-03 19:21:28 and by audio I mean bluetooth audio 2021-10-03 19:23:57 xaltsc: yes, I had that problem with pipewire too, very-very bad quality. changing the codec pipewire uses fixed the problem. 2021-10-03 19:26:18 hmm, I try to find anything by googling "pulseaudio blutooth codecs" the arch-linux wiki seems to recommend using pipewire. pipewire is a drop-in for pulseaudio, everything works the same, all pulse GUIs and tools continue to work. 2021-10-03 19:26:18 maybe you want to try it: https://wiki.archlinux.org/title/PipeWire 2021-10-03 19:26:25 ACTION googles pipewire 2021-10-03 19:26:40 I've only ever bothered learning sndio 2021-10-03 19:27:12 I generally hate pulse, so I might try pipewire 2021-10-03 19:28:28 xaltsc: pipewire does some great things: enable audio for flatpak-apps and screen-capturing for wayland. I haven't used it on alpine, yet. I am using it on arch. 2021-10-03 20:11:20 streampunk: so I've installed pipewire, audio works locally, however I don't see any pipewire related process in htop ??? 2021-10-03 20:11:52 (and bluetooth disconnects immediately after my phone has connected to the RPi) 2021-10-03 20:13:13 xaltsc: Check what /etc/xdg/autostart/pipewire.desktop executes and start this manually 2021-10-03 20:15:06 and if pipewire works you can add it to whatever autostart you have, .xinitrc .xprofile or if you have a desktop environment, use that autostart facility. 2021-10-03 20:15:49 /usr/libexec/pipewire-launcher 2021-10-03 20:19:01 You have to stop the pulse daemon, you can leave any Pulse-GUIs running. 2021-10-03 20:20:19 I've removed anything pulse or alsa 2021-10-03 20:21:28 > I've removed anything pulse or alsa 2021-10-03 20:21:28 that is ok too. 2021-10-03 20:22:37 When launching pipewire, I'm getting could not load mandatory module libpipewire-module-protocol-native though 2021-10-03 20:22:48 You have to enable all bluetooth codecs in /etc/pipewire/media-session.d/bluez-monitor.conf 2021-10-03 20:22:49 # Enabled A2DP codecs (default: all). 2021-10-03 20:22:49 bluez5.codecs = [ sbc aac ldac aptx aptx_hd ] 2021-10-03 20:23:31 xaltsc: my last message is not meant as an answer to "could not load" 2021-10-03 20:23:58 Yeah I could understand that, but i've fixed it already anyways 2021-10-03 20:26:21 xaltsc: I guess you have to install pipewire-alsa (for the normal soundcard) and pipewire-spa-bluez for blutooth 2021-10-03 20:30:41 streampunk: do I really need pipewire-alsa ? 2021-10-03 20:49:07 xaltsc: if you only use bluetooth you probably don't need it. 2021-10-03 20:57:28 My phone still won't stay connected to the rpi.... 2021-10-03 21:02:24 xaltsc: you are using the rpi as a "speaker"? I'm sorry, I seem to have misunderstood again. the direction of audio is phone -> speaker? 2021-10-03 21:04:46 phone -> rpi -> speaker yes 2021-10-03 21:05:16 ok. sorry. So you are running bluetoothd? 2021-10-03 21:05:39 yup 2021-10-03 21:06:21 xaltsc: the fuck you mean you removed anything regarding alsa 2021-10-03 21:06:26 pipewire depends on alsa 2021-10-03 21:06:31 roight? 2021-10-03 21:06:50 I don't think it depends on alsa 2021-10-03 21:06:56 did you just do something as blunt as apk info | rg alsa | parallel apk del {} 2021-10-03 21:07:00 it provides libalsa for applications that need that 2021-10-03 21:07:15 ikke: this is just according to my reading in the pipewire stack; it doesn't need alsa AT ALL? like alsamixer wouldn't even be around? 2021-10-03 21:07:23 jeffreybenzos[m]: you cannot explicitly remove packages that are installed as dependency 2021-10-03 21:07:30 i am hella skeptical of that claim considering that alsa is involved in the drivers 2021-10-03 21:07:33 yeah true 2021-10-03 21:07:48 also my parallel syntax is literally wrong iirc 2021-10-03 21:07:57 (you can replace it with a virtual empty package though, but please don't do that) 2021-10-03 21:08:01 lmao 2021-10-03 21:09:22 I didn't even install anything related to alsa in the first place, I just did apk del pulse* and apk del alsa* THEN apk install pipewirestuff 2021-10-03 21:09:28 apk add 2021-10-03 21:09:48 with -r after del 2021-10-03 21:10:40 I just have alsa-lib installed right now 2021-10-03 21:10:53 (and I have sound locally without problems 2021-10-03 21:15:24 xaltsc: I'm sorry I was thinking "raspi -> speaker". So in the other direction you are probably using https://github.com/Arkq/bluez-alsa? 2021-10-03 21:16:08 And it is probably not a audio problem but a bluetooth problem. then /etc/bluetooth/main.conf is relevant. 2021-10-03 21:18:44 So I **do** need alsa in the end ? (I'm completely lost :D) 2021-10-03 21:21:39 xaltsc: Yes, bluez-alsa, but that seems to be working, you get some audio, right? Without bluez-alsa bluetoothd could not play any audio. 2021-10-03 21:22:14 I can't even connet right now... 2021-10-03 21:22:29 The audio I got was when I was using pulse 2021-10-03 21:24:13 xaltsc: In theory bluetoothd cloud use pulse. (Now that I understand the direction: phone -> raspi). But I don't see a bluez-pulse backend, just a bluez-alsa backend. 2021-10-03 21:24:33 s/cloud/could/ 2021-10-03 21:28:10 If you stop everything that is using audio (pulse, pipewire) bluez-alsa can have exclusive control of the soundcard. And you should get audio again. But the same thing I said when I meant you are doing "raspi -> phone" is true: You should connect using the latest and greatest bluetooth audio protocols. 2021-10-03 21:29:06 I have no idea if that is something you can configure or if it is already using the best codecs/protocols. 2021-10-03 21:31:43 xaltsc: have you seen this? https://www.raspberrypi.org/forums/viewtopic.php?t=235519 2021-10-03 21:33:35 xaltsc: I'm off, let me know if you found a solution. 2021-10-03 21:36:19 yup, that's how I got it working with pulse in the first place 2021-10-03 21:36:38 s/raspi - > phone/raspi - > speaker/ 2021-10-03 21:37:20 phone --- bt ----> raspi -> speaker 2021-10-03 21:38:06 > Your system won't be seen as an A2DP capable device until pulseaudio is launched. Let's start pulseaudio on boot. 2021-10-03 21:38:06 ok at least that page says it is using pulse. 2021-10-03 21:39:11 yup, but the sound was really bad* 2021-10-03 21:39:27 so what you have is "phone > bt > raspi > pulse > alsa > soundcard > speaker", so "bt, pulse, alsa and soundcard" can be the problem. 2021-10-03 21:41:07 from the horse's mouth, ikke, pipewire doesn't fully replace alsa. it's probably confusing because of how big and disparate alsa as a concept is: since it was a whole suite, there are a lot of non-coupled parts 2021-10-03 21:41:27 what I had, now I have bluez-alsa -> ?? or -> alsa -> pipewire and it doesnt work 2021-10-03 21:42:08 maybe pulse -> pipewire will work better ? 2021-10-03 21:42:59 xaltsc: have you seen this: https://wiki.archlinux.org/title/PulseAudio/Troubleshooting#Audio_quality ? 2021-10-03 21:43:53 Local sound worked very well, how can this be relevant ? 2021-10-03 21:43:59 xaltsc: yes, pipewire and bluez-alsa (using alsa directly) could help. 2021-10-03 21:46:25 > Local sound worked very well, how can this be relevant ? 2021-10-03 21:46:25 Well it can be relevant, because its is not exactly the code. bluez can do something that is detrimental to real-time audio. But you're right it is not likely. 2021-10-03 21:46:53 ....it is not exactly the SAME code.... 2021-10-03 21:47:36 jeffreybenzos: yes, pipewire and pulse both use alsa. alsa is going to stay. 2021-10-03 21:49:14 xaltsc: and it could be a bitrate mismatch too. so what I said about asound.conf could help too. But a problem in AD2P is most likely. 2021-10-03 21:49:27 and now I am really off :-) 2021-10-03 21:56:24 as I am....... good night :) 2021-10-03 22:08:18 Someone recommended to me to install Tex live straight from the website instead of from the package repo. 2021-10-03 22:08:51 But the texlive in our repo is pretty new (march, right?); does newness matter most, here? 2021-10-04 00:29:04 > Saijin_Naib: I use pipewire as a daily driver and it works great. I don't know anything about CPU usage of it though 2021-10-04 00:29:04 Thank you! Just scrounging for more battery life when unplugged, is all 2021-10-04 00:32:10 > that's very cool! I've always thought of Alpine like a super minimal system not really suited for desktops but now I might try to create a desktop system with Alpine 2021-10-04 00:32:10 had is temporary breakage under Alpine Edge, but overall, not bad at all. Most things are just me needing extra help getting things setup as I'd like. 2021-10-04 00:32:10 I've been using Linux on and off since Ubuntu 06.10, but I'm basically user-poweruser competency, I guess. Alpine is just a really comfortable setup. APK is phenomenal. Busybox underpinning everything just feels neater. Less cognitive load starting with like, 11mb and like 100-some packages with very light dependency webs. No issues with broken debs (debsums who?), nothing getting borked during upgrades... About the worst thing I've 2021-10-04 00:33:24 I wish there was a graphical installer/boot USB so I could attract others to it, but first boot is intimidating being console-only though it is like 6 very easy setup steps and away you go. 2021-10-04 00:34:18 The wiki is great for getting you setup with a fully-functional DE, as (at least for me) that is a very new workflow coming from curses/graphical installers that go directly to full DE without me "constructing" it 2021-10-04 00:57:47 hello alpine 2021-10-04 01:24:10 Saijin_Naib[m], tiling window managers are something to look into 2021-10-04 01:24:34 drink the minimalist kool-aid 2021-10-04 03:07:17 minimalist koolaid is just water and sugar 2021-10-04 03:08:19 my favorite is seeing things like 'linux cant be minimal because of the kernel' 2021-10-04 03:08:23 always gets me a good laugh 2021-10-04 03:46:28 psykose: that's basically the reason why it can be minimal in the first place 2021-10-04 05:21:03 I'm kinda confused about what the difference is between a branch and a repository. Can a package in a repository be in multiple branches for example? 2021-10-04 05:22:17 Guest544: a branch represents separate versions of Alpine Linux 2021-10-04 05:26:16 so for example v3.14.0 was once in the testing branch, but then ended up in the main branch? 2021-10-04 05:26:53 (and subsequently got released marked as stable?) 2021-10-04 05:27:49 edge is our 'testing' branch.. Twice a year, we branch off edge (master branch in git) for stable releases 2021-10-04 05:28:04 There is also a testing repository, but that is only available for edge 2021-10-04 05:28:53 so are these branches then or repositories? https://wiki.alpinelinux.org/wiki/Aports_tree#Directories 2021-10-04 05:31:59 https://dl-cdn.alpinelinux.org/alpine/edge/ 2021-10-04 05:32:03 this is the edge repository 2021-10-04 05:32:07 edge branch* 2021-10-04 05:32:15 there you see 3 repositories 2021-10-04 05:32:22 main, community and testing 2021-10-04 05:32:44 This is the v3.14 branch: https://dl-cdn.alpinelinux.org/alpine/v3.14/ 2021-10-04 05:42:38 oh I think I'm starting to get it... main, community and testing are 3 repositories in which packages are built from (by the aports build system?) and the testing is only available for the edge branch of Alpine. Am I close? 2021-10-04 05:43:02 yes 2021-10-04 05:45:59 cool, thanks :) 2021-10-04 10:49:33 Hey guys. is it ok to install Alpine linux in USB with `setup-disk -m sys /dev/sdb`? 2021-10-04 10:58:51 Yonle: it can be ok if it's what you want :) 2021-10-04 11:01:39 Yonle: the question is, is your usb hardware suitable to run a distribution on? 2021-10-04 11:02:44 is your usb a usb thumb drive? one of those cheap ones? in the case i would not advise it. 2021-10-04 11:03:53 in this setup it would be better to run it in tmpfs mode.... and then (s)he quit... 2021-10-04 11:04:12 :\ 2021-10-04 11:04:48 i want my 3 minutes of my life back :) 2021-10-04 11:05:14 hehe 2021-10-04 11:10:04 anyone knows some cool calculator for terminal? maybe something that integrates with dmenu... 2021-10-04 11:10:28 I'm using gnome-calculator and it sometimes removes my cursor :| 2021-10-04 12:42:52 anyone know why brave browser is generally not in repos? 2021-10-04 12:44:46 what does generally mean in this context? 2021-10-04 12:45:11 i dont remember it was ever part of any repo in alpine 2021-10-04 12:47:02 i guess i havent checked much, but i dont thin its in eg void repo either 2021-10-04 12:47:16 I dont expect anybody to take up that burden, but they are free to prove me wrong. 2021-10-04 12:47:33 its just not that popular? 2021-10-04 12:48:39 i dont know its popularity, i just know its not easy to maintain a browser in a distribution, so probably not popular enough. 2021-10-04 12:49:24 ah i see 2021-10-04 12:52:56 but why would u use a browser that snuck ref codes into some urls 2021-10-04 13:07:20 D: 2021-10-04 13:07:22 ACTION hides 2021-10-04 13:11:11 u use perfect browser? also it has a nice logo 2021-10-04 13:11:32 i only use it on android actually, was just thinking to try it 2021-10-04 13:11:39 im kinda addicted to qutebrowser tho 2021-10-04 13:11:44 also trying surf 2021-10-04 13:11:45 what's special about qutebrowser? 2021-10-04 13:12:03 hints and it can talk to other applications 2021-10-04 13:12:08 hints? 2021-10-04 13:12:12 commandline 2021-10-04 13:12:23 you can press links with your keyboard 2021-10-04 13:13:19 part of why I don't use 'alternative' browsers as soon as they're powerful enough for a lot of modern pages is that they seem to lack the same features for selective blocking of JavaScript and cookies and such :| 2021-10-04 13:13:28 i dont like all the python dependencies tho 2021-10-04 13:13:48 hodapp: qb has that! 2021-10-04 13:14:27 has it a NoScript counterpart? 2021-10-04 13:14:55 i dunno noscript well 2021-10-04 13:15:03 afaik it doesnt support well plugins tho 2021-10-04 13:15:07 greasemonkey 2021-10-04 13:15:09 it has 2021-10-04 13:15:13 and a adblock 2021-10-04 13:15:20 i just use firefox with vimnium 2021-10-04 13:15:37 Misthios: i think i used something similar before qb... but it often broke 2021-10-04 13:15:42 pentadactyl 2021-10-04 13:15:56 never heard of that 2021-10-04 13:16:11 I use the minimalist browsers once in awhile, even eww in Emacs sometimes 2021-10-04 13:17:27 ew 2021-10-04 13:17:34 :P 2021-10-04 13:18:05 actually im trying to transition to rss and some script to convert html 2 markdown mostly 2021-10-04 13:18:34 but need for bank at least i guess 2021-10-04 13:19:36 emacs is good im sure, im just not a fan ;P 2021-10-04 13:19:51 has a lot of cool stuff tho 2021-10-04 13:20:22 what are you using RSS for? 2021-10-04 13:20:33 I just set up Miniflux (switched from TinyTinyRSS) and was quite impressed 2021-10-04 13:21:21 have also used Wallabag awhile (also self-hosted) and had quite good luck 2021-10-04 13:21:23 get news, some include the article in the rss feed 2021-10-04 13:21:35 or you can follow alpine rss 2021-10-04 13:22:34 ah, yeah, I use it for some news. I use FeedMe from my phone, pointed to Miniflux via the Fever API 2021-10-04 13:43:54 and you can follow video/podcasts etc, pretty much anything u can do in a browser i guess 2021-10-04 13:44:12 just less distractions and manipulation and tracking(?m 2021-10-04 13:44:34 I know Edge isn't really for daily driving, but does anyone here use it for such? 2021-10-04 13:45:05 > Though Poland is a member of the European Union, nearly 60% of Poles are strongly against replacing the złoty with the euro. 2021-10-04 13:45:09 meh, wrong chan, sorry. 2021-10-04 13:46:57 polcan is in eu 2021-10-04 13:47:01 *poland 2021-10-04 13:54:30 "I know Edge isn't really for..." <- I use `edge` on my dev vm 2021-10-04 13:55:27 Yeah, I'm thinking about hopping to Void and using Edge in a VM. 2021-10-04 13:56:16 just to be clear, you mean alpine on edge branch? 2021-10-04 13:56:29 Yes 2021-10-04 13:57:28 I found Void to be a "scam" 2021-10-04 13:58:02 Has DE on live CD, but when installed, it must be manually configured >:( 2021-10-04 13:58:39 Lmao, I mean, idc about manual configuration much if I've gone through that for Alpine. 2021-10-04 13:59:01 also the symlinking services thing is kind of annoying 2021-10-04 13:59:22 I'm still a Linux noob so idk what that is or why it's a problem. 2021-10-04 13:59:49 to enable any service, you have to make a shortcut in one place pointing to the original file 2021-10-04 14:00:13 Ahh, yeahhh, I've had to do that in the installation for a few things. 2021-10-04 14:00:13 https://docs.voidlinux.org/config/services/index.html#enabling-services 2021-10-04 14:00:15 It is kinda annoyihng. 2021-10-04 14:00:27 s/annoyihng/annoying/ 2021-10-04 14:01:44 Alpine with Plasma works good enough for me, but I might keep Void on some VM in future 2021-10-04 14:02:06 and I never had issues even when running on edge 2021-10-04 14:02:07 I'm using Alpine with Plasma, and it mostly works. 2021-10-04 14:02:19 Tbh, the only issues I have may just be KDE-related. 2021-10-04 14:02:44 or should I say, that the only problems I had were not related to me running on edge 2021-10-04 14:03:24 My main reason for wanting to try Void atm is just because I want a rolling model. 2021-10-04 14:03:39 Although I guess also because I would imagine that things are a tad more stable on that? Idk. 2021-10-04 14:03:49 though I forced myself to use stable release and only install required software from edeg 2021-10-04 14:04:21 M40e[m]: it definitely has some better support if you want desktop 2021-10-04 14:04:45 Yeahhh 2021-10-04 14:12:18 mixing stable and edge is less stable than using just edge =) 2021-10-04 14:12:36 ^ 2021-10-04 14:13:56 "Saijin_Naib, tiling window..." <- Honestly, the biggest power drain (aside from backlight but there is nothing I can do about that) is web browsing. Maybe I should just boot to console and use Lynx 🤷 No, not happening 2021-10-04 14:14:23 Hardware acceleration helps a bit with power usage 2021-10-04 14:16:32 "I know Edge isn't really for..." <- Yes. 2021-10-04 14:17:40 For how long have you been doing that? 2021-10-04 14:17:54 Just under a year at this point, across three different platforms 2021-10-04 14:18:06 Been using Alpine Edge in WSL2 for like, 2+ 2021-10-04 14:18:21 But in-hardware only OS, daily, as I said above (just under a year) 2021-10-04 14:19:09 Misthios: Yes! And I _think_ I have that working, but decisively knowing is proving difficult. StackExchange to the rescue, I guess 2021-10-04 14:19:50 edge isnt that bad 2021-10-04 14:20:24 That's a relief. I'm just looking for something rolling but I feel like I don't know much yet to tackle "everything" that might happen with using edge. 2021-10-04 14:21:27 Namely, I can't afford to have openvpn bug out, although I suppose that will be less likely to happen than with other programs/packages. 2021-10-04 14:21:36 M40e[m]: Aside from the recent kerfluffle with OpenSSL 1.1/OpenSSL 3.0 going on in Edge, I've had no issue except GDM being a jerk 2021-10-04 14:22:12 And what that needed from me was to be paitient for about a day for things to get rebuilt/patched/pushed out and then re-run my update script 2021-10-04 14:22:41 I was using Solus for two years before going to Alpine since I too wanted an "easy" rolling release 2021-10-04 14:23:08 My experience with Alpine Edge has been no worse than Solus on Shannon 2021-10-04 14:23:14 So, take that as you will 2021-10-04 14:23:33 (^ https://getsol.us/home/ ) 2021-10-04 14:24:17 jvoisin: I really like Budgie, but needing flatpaks of everything GNOME since they refuse to let anything using libhandy/libadwaita into their repos got really freaking tiresome and just untenable on 64GB eMMC 2021-10-04 14:25:03 why refuse those? 2021-10-04 14:25:41 The main developer (if not other core contributors as well) have an ideological aversion to them since they aren't desktop-first libraries and they therefore believe they are unsuitable for useage 2021-10-04 14:26:02 Nevermind that dynamic UIs that comfortable scale for screens like my 1368x768 laptop are absolutely wonderful for ME 2021-10-04 14:26:19 So, yeah. I got tired of it and finally went all-in on Alpine as a DE 2021-10-04 14:26:30 I miss snap support and that's about it 2021-10-04 14:26:33 And budgie, but ah well 2021-10-04 14:26:49 unhappy with xfce? 2021-10-04 14:27:04 id exclude gnome stuff personally but thats me :p 2021-10-04 14:27:09 No, not at all. XFCE was my first DE (minus a small stint with GNOME 2.x) 2021-10-04 14:27:18 I just really like the convenience of gnome-control-center 2021-10-04 14:27:34 And I feel like my XFCE Control Panel is missing a lot of what gnome-control-center manages for you in the GUI 2021-10-04 14:27:50 Word, I'll give edge a go for now since I can't get Void working anyway, lol. 2021-10-04 14:27:54 "mixing stable and edge is less..." <- not really 2021-10-04 14:28:21 most things on my machines are compiled statically 2021-10-04 14:28:43 eris[m]: I really like their stuff, haha. Familiarity? I don't know. A lot of their HIG makes sense to me, I think 2021-10-04 14:30:31 jvoisin: I would love LXDE from Arch with their GTK3/Wayland patches, personally. I'm trying and failing to get them built 2021-10-04 14:30:32 gnome is just more annoying as a packager 2021-10-04 14:30:57 Oh, interesting. Can you noobsplain that for me? 2021-10-04 14:31:03 KDE is kinda annoying, but idk. 2021-10-04 14:31:14 last time i tried i got annoyed and quit 2021-10-04 14:31:18 I would think QT would be worse with their recent changes to how they're only offering support for community/free builds for not very long 2021-10-04 14:31:18 Saijin_Naib[m]: try to get those patches upstreamed then :) 2021-10-04 14:31:20 Might not be the same topic of discussion though, lol. 2021-10-04 14:31:24 kde is hell bcos qt5 takes ages 2021-10-04 14:31:42 M40e[m]: It is overwhelming for me. And runs awful on my laptop 2021-10-04 14:32:01 jvoisin: If you knew how bad I was at all this, you would know that's not happening haha 2021-10-04 14:33:08 I think it works fine for the most part, but I'm running into too many smaller issues, so idk. I'm looking to swap sddm for some other DM, but there are more things to learn about first. 2021-10-04 14:33:23 sway 2021-10-04 14:33:33 and beautiful crashing 2021-10-04 14:33:38 Yeah, that's what I was considering. 2021-10-04 14:33:46 Lol, I don't want any crashes though. 2021-10-04 14:34:04 don't worry, it only crashed for me when hovering over taskbar 2021-10-04 14:34:19 * hovering over window icon on taskbar 2021-10-04 14:35:42 Oof. 2021-10-04 14:36:09 and Wayland + `sddm` was just a black screen 2021-10-04 14:40:07 For me X11 and sddm is a black screen. 2021-10-04 14:40:38 But I couldn't seem to find how to map input from an external touch display under Wayland. 2021-10-04 14:40:40 Nice! GDM for me is broken. I hope the GNOME 41 stack upgrade fixes that 2021-10-04 14:40:42 C'est la vie. 2021-10-04 14:40:49 lightdm seems to play nice though 2021-10-04 14:41:32 Does Wayland need the xf86-input-libinput for that maybe? 2021-10-04 14:42:05 Maybe. I'll try installing that and see how it fares. Anyway, off to work. Deuces. 2021-10-04 14:57:39 Saijin_Naib[m]: yep it broke strangely some days ago, could you check your '/var/lib/gdm/.local/share/xorg/Xorg.0.log'? I see: "[ 120.335] (EE) xf86OpenConsole: Cannot open virtual console 1 (Permission denied)" 2021-10-04 14:58:11 I tried changin permissions but got another error 2021-10-04 15:00:30 donoban: Sad/glad it wasn't just me. I am also having really strange issues under GDM where it won't re-invoke itself after sleep/resume or any display DPMS. I have to bring it back by switching to a VT and then back to GDM, or sometimes just force reboot. LightDM behaves way better in this regard 2021-10-04 15:06:14 uhM 2021-10-04 15:06:30 well maybe that log helps you digging into it 2021-10-04 15:49:11 Has anyone tried installing the nVidia proprietary drivers under Alpine? I see that Void provides them as a package 2021-10-04 15:49:31 I know we have noveau, but that isn't really tenable for my GPU since it'll be stuck in one p-state :\ 2021-10-04 15:49:40 void is a glibc distro 2021-10-04 15:49:57 pretty sure nvidia drivers need glibc for some dumb stupid reason 2021-10-04 15:50:03 some obscure function 2021-10-04 15:50:04 eris[m]: They provide it under their MUSL-one as well, unless I read the packages wrong 2021-10-04 15:51:34 nvidia says 'fuck you' 2021-10-04 15:51:36 https://github.com/NVIDIA/nvidia-installer/issues/10 2021-10-04 18:54:28 Anyone willing to hand-hold me through getting lxmusic built for aports? 2021-10-04 18:55:29 I want all of LXDE, but I doubt I can afford what it would cost to contract someone to do it for me 2021-10-04 18:55:50 (specifically the neat GTK3/Wayland patched version from Arch for forward-compat) 2021-10-04 19:02:49 I'm just playing whack-a-mole with dependencies from it's configure script 2021-10-04 19:02:57 Does this count as development? 2021-10-04 19:03:30 Does for me :) 2021-10-04 19:03:45 Is there a smarter or correct way to do this? 2021-10-04 19:03:49 I'm 100% sure I'm wrong here haha 2021-10-04 19:03:57 Looking for a readme? 2021-10-04 19:04:00 otherwise, not really 2021-10-04 19:04:19 Their readme is... pretty light 2021-10-04 19:04:33 I'm trying to write down each package I installed taht made something go from red/yellow to green in the configure script 2021-10-04 19:04:45 Since I think I need them for the abuild script 2021-10-04 19:05:00 Starting from a naked Alpine WSL2 image which only had alpine-sdk 2021-10-04 19:15:12 Do I need to worry about the libraries pulled in by the dependency I installed for the abuild script? 2021-10-04 19:15:17 Or just let abuild handle them? 2021-10-04 19:16:32 And add them as needed? 2021-10-04 19:16:49 for instance, I added libmodplug-dev, and it pulled in libmodplug 2021-10-04 19:16:53 You typically add the *-dev packages as makedepends in your APKBUILD 2021-10-04 19:16:55 yes 2021-10-04 19:17:00 I don't need to write both in? 2021-10-04 19:17:01 that's good 2021-10-04 19:17:03 no 2021-10-04 19:17:22 abuild traces soname dependencies 2021-10-04 19:18:37 I was thinking lxmusic would be an easy build, but it needs xmm2s which is... This is nuts haha 2021-10-04 19:20:17 Yeah, sometimes it goes like that 2021-10-04 19:20:58 This is why ya'll get paid more than I do 2021-10-04 19:21:00 Fuck this haha 2021-10-04 19:22:59 It takes endurance 2021-10-04 19:23:28 Seriously. This is maddening. Everything I plug up reveals like 30 other things I'm missing 2021-10-04 19:23:36 and I'm not sure our repos even have the stuff I'm missing 2021-10-04 19:23:43 Which means what? I have to build those too!? 2021-10-04 19:24:22 yup 2021-10-04 19:25:59 So this whole process for each of those to get them in aports first 2021-10-04 19:26:32 Saijin_Naib[m]: It can all be added in one go 2021-10-04 19:26:54 You can first create packages for them locally so that you can get the target package to build 2021-10-04 19:26:59 and then make a single MR to add them 2021-10-04 19:27:37 hmm 2021-10-04 19:27:48 But I still need to make the apkbuild thing correct for each 2021-10-04 19:27:51 Regardless 2021-10-04 19:27:55 So more whack-a-mole 2021-10-04 19:29:03 :) 2021-10-04 19:49:46 do the dev packages not install .h files? 2021-10-04 19:49:51 I have ffmpeg-dev installed but the configure script is complaining 2021-10-04 19:49:52 Checking for header avcodec.h 2021-10-04 19:51:30 they do install headers: https://pkgs.alpinelinux.org/contents?file=avcodec.h 2021-10-04 19:52:27 would've been strange if they didn't 2021-10-04 19:52:44 Checking for 'libavcodec' : yes... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/HWjfxlUadNPoaPpCkxHZNyJN) 2021-10-04 19:53:07 Maybe the script is checking the wrong place or soemthing? 2021-10-04 19:53:54 Oh... Okay 2021-10-04 19:54:01 it finds libavcodec/avcodec.h though 2021-10-04 19:54:03 I'm trying to get everything green. That's the level I'm on 2021-10-04 19:54:04 So 2021-10-04 19:54:09 Yeah 2021-10-04 19:55:11 So is this the same thing then?... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/oQBNBwDQrluZpsmmEuocCASJ) 2021-10-04 19:55:36 seems like it 2021-10-04 19:55:50 i mean, if it builds, then it should be fine 2021-10-04 19:56:24 your configure script simply checks for all possible places where a header could be and reports that apparently 2021-10-04 20:04:29 Wump. Won't compile. Time to tap out, I guess 2021-10-04 23:35:09 was just messing around with the `5.10.61-0-virt` kernel, and I noticed that it doesn't have support for the Ensoniq ES1371/ES1373 / Creative Labs CT2518 sound card 2021-10-04 23:35:31 that sound card is the standard one for all VMware VMs 2021-10-04 23:37:10 I'm assuming for this particular VM that I would need to switch the kernel out? Is it easy to rebuild alpine kernels with additional modules? 2021-10-05 00:08:47 I managed to figure it out, so that's a good sign haha 2021-10-05 00:08:57 Do you have abuild setup? 2021-10-05 00:09:14 I had to do it to make a MR to get my touchpad supported in the regular LTS kernel 2021-10-05 01:32:12 Good night, I'm going crazy with my wifi, it doesn't work at all, neither with iwd nor wpa_supplicant, with networkmanager. 2021-10-05 01:34:53 When I click on enable wifi it says enabled but not enabled. And the message in /var/log/messages. 2021-10-05 01:34:55 daemon.info NetworkManager[2607]: [1633397674.6293] audit: op="radio-control" arg="wireless-enabled" pid=2934 uid=1000 result="success" 2021-10-05 01:35:37 with iwd 2021-10-05 01:38:26 But the SSID doesn't appear to click 2021-10-05 01:39:59 how do I change the kernel cmdline options 2021-10-05 01:40:53 With wap_supplicant the SSID appears but it does not connect. I've already changed the firmware or driver, there are times when it appears and others not, very tense. 2021-10-05 03:03:21 "how do I change the kernel..." <- I edit the file in /etc/default/grub 2021-10-05 03:03:34 Then update the grub config 2021-10-05 03:03:51 "With wap_supplicant the SSID..." <- What services do you have in your rc-update? 2021-10-05 03:04:26 iwd is incredibly hit-or-miss with me as well. I'm fallen all the way back to just wpa_supplicant+networking, not even NetworkManager since that is being a pain as well 2021-10-05 03:28:11 hey, are there any issues with the server that setup-alpine connects to? i'm getting wget operation timed out trying to load mirrors 2021-10-05 03:29:39 are you sure the network works 2021-10-05 03:30:29 yes, i can ping my local alpine mirror 2021-10-05 03:31:25 can you wget a website 2021-10-05 03:31:55 yes 2021-10-05 03:32:17 hm, i can wget google but not alpinelinux.org 2021-10-05 03:32:36 sounds like dns 2021-10-05 03:34:20 i can dig alpinelinux.org if i install bind-tools 2021-10-05 03:34:35 seems to work 2021-10-05 03:34:36 can you ping alpinelinux.org? 2021-10-05 03:35:07 nope 2021-10-05 03:37:09 that is strange 2021-10-05 03:39:58 yeah, i really have no idea what's happening here 2021-10-05 03:40:16 maybe the connection thing picks ipv6 for alpinelinux.org and your ipv6 is broken 2021-10-05 03:40:48 i've had that before for random domains since the router gave out ipv6 but couldn't actually route it 2021-10-05 03:41:00 i indeed do not have ipv6, but if i dig the ipv4 and ping that directly, it doesn't work either 2021-10-05 03:41:09 bizarre 2021-10-05 03:41:43 does setup-alpine let me specify a mirror manually? it seems that it just makes me pick from an (inexistent) list 2021-10-05 03:47:28 ohh, it lets me edit the text file directly, nice 2021-10-05 03:47:35 let's see if this install actually works 2021-10-05 04:01:49 welp, i talked to our netadmin and my IP was configured incorrectly on the router 2021-10-05 04:01:59 all working now :) 2021-10-05 04:51:34 oh lol 2021-10-05 05:01:22 looks like as of last january snort3 is considered stable, any thoughts on adding it to edge? 2021-10-05 05:05:28 Entelin: so you mean to upgrade snort? 2021-10-05 05:12:46 looks like snort2 is in the repo but not snort3. I'm just getting started with snort in general so I feel like it makes more sense to learn 3 since thats evidently the future. 2021-10-05 08:35:48 hi. is anybody out there who can tell me how to fix connect to lets encrypt secured webservers in alpine 3.8? regarding the root-ca change from september 30th 2021-10-05 08:36:38 i tried removing the "old" root-ca like here https://www.openssl.org/blog/blog/2021/09/13/LetsEncryptRootCertExpire/ 2021-10-05 08:36:55 but still not getting connection via https 2021-10-05 08:37:02 update to a newer alpine version, of course 2021-10-05 08:37:27 yeah i tried. my image is based on nginx:mainline-alpine 2021-10-05 08:37:32 and this is using 3.8 2021-10-05 08:37:46 docker? 2021-10-05 08:37:49 yes 2021-10-05 08:37:53 official nginx? just pull the new image 2021-10-05 08:38:33 latest nginx:mainline-alpine is on 3.14 2021-10-05 08:39:46 hm ok i will try again 2021-10-05 08:40:45 i don't remember if docker ever re-pulls images it has in cache, and if not it means you never manually did a pull on it for like 3 years 2021-10-05 08:41:04 iy fordn'y 2021-10-05 08:41:12 * it doesn't 2021-10-05 08:41:41 or, lost your cache/changed pcs in 3 years too :) 2021-10-05 08:42:08 yes you're right. using --no-cache it takes 3.14 2021-10-05 08:42:18 i've never seen someone actually run into this, but it's fixed with just a docker pull nginx:mainline-alpine and then rebuilding the container 2021-10-05 08:42:29 or no cache, yeah 2021-10-05 08:42:57 normally, you should be using something like `compose` file or watchtower 2021-10-05 08:43:17 i don't think compose changes this behaviour 2021-10-05 08:43:27 you can if you set option to always pull 2021-10-05 08:45:43 or it could have been a custom fix I did for myself 🤔 2021-10-05 08:45:51 but that probably means that you're doing something else wrong in your deployment process 2021-10-05 08:46:42 if you're deploying to prod, image tags should be unique and (almost always) refer to the same artifacts 2021-10-05 08:46:51 ie. using :latest in prod is a bad idea 2021-10-05 08:47:03 true, you should also have reproducible images ;) 2021-10-05 08:47:40 there might be some cases where using semantically versioned/tagged releases might make sense to AlwaysPull 2021-10-05 08:48:32 but if you're doing anything like that at all, I hope your deployment process is smart enough to roll out slow and small, and auto rollback on failure 2021-10-05 08:50:01 as always, proper testing and use of canaries can mitigate issues there :) 2021-10-05 14:39:31 is /usr/lib/qt6/bin/ on the path? If so, why can't Alpine edge find it for calling qmake? I stat'd /usr/lib/qt6/bin/qmake and it is there/reachable and if I invoke it directly, qmake will run, but I can't just "qmake" from terminal 2021-10-05 14:43:32 echo $PATH and see if it's there 2021-10-05 14:45:24 It is not... Hmm 2021-10-05 14:45:30 Looks like that is consistent with QT5 dev 2021-10-05 14:45:48 But from looking at build instructions for other platforms, we might be doing something different in that regard to having it not on path 2021-10-05 14:46:35 normally i believe someone adds a /etc/profile.d/xyz entry for it and a profile source will make the path correct 2021-10-05 14:46:45 it appears the package just doesn't ship with that 2021-10-05 14:47:01 Okay, cool. Is that wrong, right, or kind of neither 2021-10-05 14:47:10 Like, should I file an issue in aports? 2021-10-05 14:47:17 i dunno really 2021-10-05 14:47:25 neither 2021-10-05 14:47:43 on gentoo at least those kinds of things do get added to profile 2021-10-05 14:47:49 here you can just manually add it to the shell profile 2021-10-05 14:47:52 it's just a line anyway 2021-10-05 14:48:05 Linux is just terrible when it comes to environment variables 2021-10-05 14:48:08 Hello, Is the mailinglist-bot dead for aports patches atm? A patch i sent hasnt shown up yet 2021-10-05 14:48:35 probably better to ask in #alpine-devel 2021-10-05 14:49:36 psykose: But it is a barrier for low-information folks like myself 2021-10-05 14:49:46 But yeah, now that I know what it is, I can fix it 2021-10-05 14:50:00 I have a .sh in my profile.d for aliases and whatnot 2021-10-05 14:50:05 i don't disagree :) 2021-10-05 14:50:52 Imagine if there was some kind of file that could set global envvars on Linux 2021-10-05 14:51:08 If this is a joke I don't get it haha 2021-10-05 14:51:13 Is there a thing? 2021-10-05 14:51:32 well, there is `/etc/environment` as long as you follow the logon process 2021-10-05 14:51:45 because `pam_env` is 'wonderful' thing 2021-10-05 15:05:26 Hmm... If something builds against qt5, is that pretty much a dead-end for inclusion in abuild? 2021-10-05 15:05:34 Since we have qt6 now? 2021-10-05 15:08:17 both exist 2021-10-05 15:09:08 No qt5 is fine 2021-10-05 15:09:11 There is nothing in aports that use qt6 right now so blocking anything on qt5 would be stupid 😛 2021-10-05 15:13:14 Thanks for helping me, all haha 2021-10-05 15:13:29 https://github.com/nickbnf/glogg 2021-10-05 15:13:29 Its for this, btw: 2021-10-05 15:13:37 Once I get that sorted, I want to try klogg 2021-10-05 15:13:47 https://github.com/variar/klogg 2021-10-05 15:14:14 glogg saved my ass when I worked for a precision ag. company. Custom highlighting and filtering of logs with real-time file watch/refresh is amazing 2021-10-05 15:14:52 I think I have to give up on xmms2/lxmusic since I'm getting errors that seem to be C++ errors, not just missing dependencies from whack-a-mole :( 2021-10-05 15:15:00 what errors 2021-10-05 15:15:31 Oh jeez... I should have saved it. Something about an undefined reference to another file 2021-10-05 15:15:41 on an include 2021-10-05 15:15:43 i'd also start with klogg because it seems to be the maintained one 2021-10-05 15:15:48 and has the same build setup more or less 2021-10-05 15:16:17 Oh, okay. That makes sense. I was thining glogg might be safer for me since it is more or less static and I don't know how I'd even begin to approach backports or fixing 2021-10-05 15:17:52 And I should be using Alpine Edge, right? 2021-10-05 15:17:56 Or latest-stable 2021-10-05 15:20:08 for making new packages? technically you should build on edge 2021-10-05 15:20:23 as they will only be available there and will build with edge deps 2021-10-05 15:21:04 you don't need to be on edge to do that technically, you can just keep an edge chroot around 2021-10-05 15:21:12 but the simplest way is to just use edge, no effort 2021-10-05 15:22:31 That works for me 2021-10-05 15:22:42 A few seconds to reset my WLS2 and go fresh if I mess something up again 2021-10-05 15:23:11 if you use wsl you can just have multiple installs there too 2021-10-05 15:23:28 (wsl2 because I loathe building on 64gb eMMC, 4GB RAM, and Intel N3450 laptop) 2021-10-05 15:24:08 Yeah, that's true. Tempted to setup a "clean" baseline rootfs tar I can just import/export when needed that has alpine-sdk, doas 2021-10-05 15:24:13 Save me like, 30 seconds of work haha 2021-10-05 15:24:35 the alpine chroot script basically does exactly that 2021-10-05 15:24:50 -_- Sigh I know nothing 2021-10-05 15:25:02 So I don't have to nuke my whole instance every time I want to try to build something else? 2021-10-05 15:25:03 noo don't feel that way 2021-10-05 15:25:13 you don't have to, no 2021-10-05 15:25:26 you can use the --virtual adds with apk to be able to easily clean them 2021-10-05 15:25:33 echo 'alpine-sdk' >> /etc/apk/world 2021-10-05 15:25:33 apk fix 2021-10-05 15:25:33 or use abuild rootbld 2021-10-05 15:25:42 or use the builds in docker 2021-10-05 15:25:50 there's probably like 10 ways to manage that 2021-10-05 15:25:59 endless possibilities ;) 2021-10-05 15:26:46 Oh god I'm lost 2021-10-05 15:26:47 PJ, what does your thing do 2021-10-05 15:27:04 Oh, is it also possible to like... "lock" certain pakages so I can just do an apk del * and not nuke the 11 packages it needs to keep running haha 2021-10-05 15:28:01 Saijin_Naib[m]: it installs the package 2021-10-05 15:28:22 although it's not automatic nor does it persists reinstall 2021-10-05 15:28:46 Does anyone do contract builds for aports? haha 2021-10-05 15:29:07 you should just add abuild-rootbld and use that to build your apkbuilds 2021-10-05 15:29:21 psykose: Is that the chroot you mentioned? 2021-10-05 15:29:26 that or docker 2021-10-05 15:29:29 it's the automatic version of it 2021-10-05 15:29:39 it just makes a chroot, adds the makedepends of the apkbuild, and runs the build in it 2021-10-05 15:29:53 + some extra stuff that doesn't matter here 2021-10-05 15:29:55 So I need to have my apkbuild script written for that to work? 2021-10-05 15:30:02 it's the same script 2021-10-05 15:30:02 "A few seconds to reset my WLS2..." <- while I'm currently trying to find where my disk is 2021-10-05 15:30:04 you don't need anything 2021-10-05 15:30:25 well, you need abuild-rootbld 2021-10-05 15:30:30 ;) 2021-10-05 15:31:05 Okay, added that to my fresh iamge 2021-10-05 15:31:12 Does that have its own wiki? 2021-10-05 15:32:28 apparently not 2021-10-05 15:32:39 `build -h | grep rootbld` 2021-10-05 15:32:58 s/build/abuild/ 2021-10-05 15:33:20 Ah, cool. 2021-10-05 15:33:27 rootbld Build package in clean chroot 2021-10-05 15:34:10 So how then do I interact with this? 2021-10-05 15:34:25 Normally, I would git clone the repository, cd into it, and try configure 2021-10-05 15:34:34 Write down what packages I need in OneNote 2021-10-05 15:34:44 and then try build when everything is green 2021-10-05 15:34:54 If it works, yay, I ahve a package list. I never got further than that 2021-10-05 15:35:05 just add them to makedepends in APKBUILD? 2021-10-05 15:35:16 and you can `build -r` 2021-10-05 15:35:23 s/build/abuild/ 2021-10-05 15:39:05 s/and you can `build -r`/and you can `abuild -r` or rather `abuild -r rootbld`/ 2021-10-05 15:44:25 Thank you 2021-10-05 15:44:38 I'm not still not getting how I need to interact with it, but I'll see what I can find 2021-10-05 15:44:58 are you making a new package or updating an old one 2021-10-05 15:46:48 https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package 2021-10-05 15:46:48 should have everything required to bootstrap your brain for making packages 2021-10-05 16:04:21 Adding new one(s) to aports 2021-10-05 16:04:25 first time 2021-10-05 16:04:43 All I've managed thus far is local compile of glogg and submitting a MR for adding a module to linux-lts 2021-10-05 16:04:50 Super, super low-skill/information here 2021-10-05 16:16:35 ah, i assumed you had already made a new one before :) that guide above should get you through the start 2021-10-05 16:23:36 Nah, not yet. 2021-10-05 16:59:26 has anyone got apk complain about maskins when building packages? 2021-10-05 17:00:02 i have a few packages from stable and rest is edge but it complains about version only available on stable 2021-10-05 17:00:13 more specifically qt5 2021-10-05 17:00:57 i don't have pinnings for those packages on either edge or stable 2021-10-05 17:02:33 i am trying to build mumble with pa support as usual but it complains a version is masked on stable 2021-10-05 17:03:18 all packages i have installed should be from edge and i cannot really afford running apk upgrade -a due to some pinnings 2021-10-05 17:25:54 Has anyone noticed thttpd throttles file has some entries ignored? I am testing on devuan but target is eventually alpine. 2021-10-05 17:36:24 Entries like 'stuff*/**.js 12345' do not work at all, for example 2021-10-05 18:25:36 ? 2021-10-05 19:18:08 ... 2021-10-05 20:20:20 Is it possible to run a MUSL-compiled executable from Void-MUSL under Alpine? I know, dependency-hell if it is dynamic(?) linked, but just as a proof of concept 2021-10-05 20:21:28 I'm struggling really hard to compile xmms2/lxmusic, and the whole point is I believe it should have a much lower CPU consumption during playback than the python based players I'm using now. Be a kick in the rear to take weeks to find out I was wrong with my assumption that it is any more efficient haha 2021-10-05 20:22:30 Saijin_Naib[m]: should do 2021-10-05 20:22:36 alpine binaries work on kiss 2021-10-05 20:25:53 Neat. Now I need to further check my assumption that Void had it packaged (it doesn't) and find someone that does haha 2021-10-06 00:24:41 I' 2021-10-06 00:24:43 oops 2021-10-06 00:26:21 I'm trying to sign up for Alpine's GitLab instance, but I'm not getting a confirmation email. 2021-10-06 08:30:01 lonjil: have you made it work? 2021-10-06 10:03:20 Anyone on thttpd 2.29 thorttles file not being executed/read right? 2021-10-06 10:11:20 throttles even 2021-10-06 10:25:31 Okay, clarifying bug/feature: a html tree dir/file called 'foo%2b/' is not accessible (visible) from web side because 'GET foo%2b' is interpreted as a url escape for 'foo+' which does not exist on the server fs. 2021-10-06 10:26:00 Morale: do not use files with url escape like sequences in their names, or they will not be visible to the web clients. 2021-10-06 10:26:12 That is actually a feature! :) 2021-10-06 11:39:09 After more research: if the global throttle setting is present in the throttle.conf file '** 10000' for example, then the other entries are not parsed right and or ignored. 2021-10-06 11:41:31 Also, %xx url escapes containing files are working, not "hidden" as I claimed. 2021-10-06 11:41:57 Anyone using thttpd on alpine really? 2021-10-06 13:41:52 Okay, fixed thttpd throttling, a source edit it needed, it is a logic error in throttle data processing. Ping me if you need a patch. It is just one line as I said. 2021-10-06 13:42:26 I'm on here as user____ sometimes and also on libera sometimes 2021-10-06 13:51:02 clandmeter: I just asked gitlab to resend the confirmation email for the 4th time, and now it worked. 2021-10-06 14:02:00 lonjil: good, i think we fixed it :) 2021-10-06 14:02:23 woo 2021-10-06 14:20:50 user____: without reading http rfc, the server logically must decode url escapes in the request. otherwise you will be unable to access any files containing spaces in the filename 2021-10-06 14:22:15 That too. But this is not about filenames, it is about a logic flaw in thttpd throttles file data application. Natively, it picks the lowest speed from all matches which apply. I made it apply the speed of the 1st match. 2021-10-06 14:37:35 ACTION uploaded an image: (29KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/QbaVatMhGWNpMOxdXYyNIuhr/image.png > 2021-10-06 14:37:53 New day, new rootfs. I'm going to try and preserve this one for abuild-rootbld like you all recommended 2021-10-06 14:38:10 My goal is to get at least one abuild recipe right 2021-10-06 15:38:26 Is Rust stuff a bad thing to try and make my first aports with? 2021-10-06 15:38:30 https://github.com/Peltoche/lsd 2021-10-06 15:38:41 nope 2021-10-06 15:38:56 I have what I think is a decent APKBUILD so far. It grabbed the source tarball properly and extracted it when I ran abuild unpack 2021-10-06 15:38:59 try looking at existing rust packages 2021-10-06 15:39:13 uhh 2021-10-06 15:39:15 I think afetch is one 2021-10-06 15:39:48 Damnit, no, it's C 2021-10-06 15:40:16 https://git.alpinelinux.org/aports/tree/community/sniffglue/APKBUILD 2021-10-06 15:40:17 https://pkgs.alpinelinux.org/package/edge/community/x86_64/dust 2021-10-06 15:40:58 Thank you both 2021-10-06 15:41:11 Sorry for the need for hand-holding 😬 2021-10-06 15:41:43 No worry 2021-10-06 15:42:11 Is that funky prepare{} step that dust is doing generic? Can that apply to lsd? 2021-10-06 15:42:20 there is also `procs`, `ripgrep`, `pijul`, `hyperfine` 2021-10-06 15:42:47 Saijin_Naib[m]: it can 2021-10-06 15:43:29 it can apply to all Rust binaries 2021-10-06 15:44:21 Is it something I should include? 2021-10-06 15:44:36 I don't understand what it is doing other than the comment, which I think fits the Alpine vibe 2021-10-06 15:44:46 But like, I don't know what the risk/reward or trade-offs for this are 2021-10-06 15:44:51 Extra compile time? I dunno 2021-10-06 15:45:06 It can increase start time by little 2021-10-06 15:45:15 hmm 2021-10-06 15:45:27 That might be annoying for something you'd alias ls for 2021-10-06 15:45:31 you can read more here: https://opensourcelibs.com/lib/min-sized-rust 2021-10-06 15:45:47 But that won't matter once it is run the first time in a session right? Doesn't linux cache stuff pretty aggressively? 2021-10-06 15:46:36 I don't think that will impact speed much 2021-10-06 15:47:06 Oh garsh here we go haha 2021-10-06 15:48:47 I would keep the optimisation for now, if there will be any issues it can be removed 2021-10-06 15:50:20 This is daunting haha 2021-10-06 15:51:53 Alright, think I got it all templted from the dust example 2021-10-06 15:52:12 Thank you, PJ/ikke 2021-10-06 15:52:18 No worry 2021-10-06 15:52:28 btw, we have a dedicated development channel (#alpine-devel) 2021-10-06 15:52:51 I know 2021-10-06 15:52:58 I do not think I'm at a level where I belong there yet 2021-10-06 15:53:01 Too beginner 2021-10-06 15:53:08 I'm reading what goes on there haha 2021-10-06 15:57:45 >>> ERROR: lsd: source 0.20.1.tar.gz needs to be renamed to avoid possible collisions 2021-10-06 15:58:40 source="$pkgname-$pkgver.tar.gz::https://..." 2021-10-06 16:01:36 Okay so that line still applies? 2021-10-06 16:01:47 I was confused because I was getting the package from github and it does have the package version 2021-10-06 16:01:58 source="https://github.com/Peltoche/$pkgname/archive/refs/tags/$pkgver.tar.gz" 2021-10-06 16:02:11 yes, but just the version, not the package name 2021-10-06 16:03:08 Worked beautifully, thank you so much! 2021-10-06 16:08:12 oh, btw regarding the optimisation you can try building the package and see how big the binary is 2021-10-06 16:08:33 I don't remember how I installed `lsd` but it's ~3MB 2021-10-06 16:08:38 I think that will be next steps. I'm getting some arcane Cargo.toml parsing error 2021-10-06 16:08:53 log? 2021-10-06 16:08:56 Yeah, LSD is mega-tiny. The dev provides musl binaries, so I used to just mv them to my bin directory 2021-10-06 16:09:10 Caused by: 2021-10-06 16:09:10 error: failed to parse manifest at `/home/alpine/Sources/aports/testing/lsd/src/Cargo.toml` 2021-10-06 16:09:10 virtual manifests must be configured with [workspace] 2021-10-06 16:09:32 I'm trying to read up on this now, but I don;t know enough to know what I should be looking for haha 2021-10-06 16:13:26 Saijin_Naib[m]: ..is 3mb for directory listing tiny?? 2021-10-06 16:13:49 I guess when you compare it to the fonts needed haha 2021-10-06 16:13:50 busybox is 800kb.... 2021-10-06 16:14:00 at least on my kiss system 2021-10-06 16:14:02 busybox doesn't have colours, runtime, etc. 2021-10-06 16:14:09 busybox has colours 2021-10-06 16:14:16 what the fuck is 'runtime' 2021-10-06 16:14:20 `time ls` 2021-10-06 16:14:23 ? 2021-10-06 16:14:28 no 2021-10-06 16:14:34 it's not `run time` 2021-10-06 16:14:37 it's `runtime` 2021-10-06 16:14:52 https://stackoverflow.com/questions/3900549/what-is-runtime 2021-10-06 16:14:56 that doesnt answer my question 2021-10-06 16:15:11 neither does a link 2021-10-06 16:15:24 do you mean.. the programming language runtime? 2021-10-06 16:15:29 sorry, but I'm currently on the clock and need to run away, both Rust and Go have runtimes included in binary so they can work 2021-10-06 16:15:37 yes 2021-10-06 16:15:44 800kb is the figure including my statically linked musl... 2021-10-06 16:16:09 also not really into arguing what is better, if you don't like 3MB `ls`, don't install it 2021-10-06 16:16:09 Maybe it'll be smaller with all the optimizations 2021-10-06 16:16:13 I have not gotten that far yet 2021-10-06 16:16:26 im not saying to install it or not 2021-10-06 16:16:34 im saying that its not small ^^' 2021-10-06 16:16:48 it is small comparing to other utilities written in Rust 2021-10-06 16:16:52 cant even fit on a damned floppy disk 2021-10-06 16:17:03 Heh, perspectives/normals are funny 2021-10-06 16:17:04 i thought rust was meant to produce small binaries? 2021-10-06 16:17:09 they are small 2021-10-06 16:17:12 I thought it was meant to produce "Safe" ones 2021-10-06 16:17:20 and safe :v 2021-10-06 16:17:35 once you use aggressive optimisations you can get really nice small bins 2021-10-06 16:17:54 or use whitney C :^) 2021-10-06 16:17:56 whitney rust 2021-10-06 16:18:08 is that a thing 2021-10-06 16:18:09 my Discord bot was 20MB with default values, it's ~6MB after opimisation 2021-10-06 16:18:14 i think rust is too verbose to do those kinds of shenanigens 2021-10-06 16:18:30 you can strip it out 2021-10-06 16:18:40 https://opensourcelibs.com/lib/min-sized-rust 2021-10-06 16:18:54 gotta run, i'll be back 2021-10-06 16:19:32 eris, you good with abuild/rust stuff? haha 2021-10-06 16:19:33 I'm stuck 2021-10-06 16:19:39 no 2021-10-06 16:19:41 i hate rust 2021-10-06 16:19:50 eris[m]: https://codeberg.org/ngn/k/src/branch/master/v.c this is whitney C 2021-10-06 16:20:08 eris[m]: rust is definitely too verbose to pull this off :p 2021-10-06 16:20:10 Oh god that's awful. Looks like brainfuck 2021-10-06 16:20:31 wait until you see what programming language this implements, ay? 2021-10-06 16:20:36 give me a programming problem 2021-10-06 16:20:44 I don't program 2021-10-06 16:20:47 Wrong person for this game 2021-10-06 16:20:47 haha 2021-10-06 16:20:52 fiine 2021-10-06 16:20:53 ill do like 2021-10-06 16:20:57 fib 2021-10-06 16:21:32 `{1_x(|+\)/!2}` 2021-10-06 16:21:37 no tacit :( 2021-10-06 16:22:20 I don't know what that means 2021-10-06 16:22:55 tacit means a function that doesnt use variables 2021-10-06 16:23:28 `sum = foldr (+) 0` tacit 2021-10-06 16:23:40 ` sum' xs = foldr (+) 0 xs` 2021-10-06 16:23:41 not tacit 2021-10-06 16:23:52 one of these has extreneous info 2021-10-06 16:23:52 the other doesnt 2021-10-06 16:26:15 hmm 2021-10-06 16:26:22 Code Golf type stuff? 2021-10-06 16:26:31 no 2021-10-06 16:26:38 just removing unneccessary info 2021-10-06 16:31:01 That's basically making use of currying, right? 2021-10-06 16:52:10 not a clue! 2021-10-06 19:51:02 "eris, you good with abuild/..." <- what you need? 2021-10-06 20:15:13 I'm still just stuck on parsing what that Rust Cargo.toml error needs to be fixed 2021-10-06 20:15:28 Caused by: 2021-10-06 20:15:28 virtual manifests must be configured with [workspace] 2021-10-06 20:15:28 error: failed to parse manifest at `/home/alpine/Sources/aports/testing/lsd/src/Cargo.toml` 2021-10-06 20:17:08 could you push it to your fork on gitlab.alpinelinux.org ? 2021-10-06 20:19:56 but I think you have some pathing issue since `Cargo.toml` shouldn't be in `src/` 2021-10-06 20:34:38 Yeah, I adjusted the path for the prepare script and now it is inserting the optimizations into the end of the Cargo.toml file, which is great, but it still fails 2021-10-06 20:34:39 >>> ERROR: lsd: build failed 2021-10-06 20:34:39 error: could not find `Cargo.toml` in `/home/alpine/Sources/aports/testing/lsd/src` or any parent directory 2021-10-06 20:34:42 Pushing now 2021-10-06 20:37:24 Saijin_Naib[m]: sounds like your builddir is incorrect 2021-10-06 20:37:44 what is the contents of src/ 2021-10-06 20:38:39 lsd-0.20.1 lsd-0.20.1.tar.gz 2021-10-06 20:38:59 and under the lsd-0.20.1 is where all the extracted source code is, and the patched Cargo.toml now 2021-10-06 20:39:07 It was ending up at this directory level before 2021-10-06 20:50:40 https://gitlab.alpinelinux.org/Saijin-Naib/aports/-/tree/master/testing/lsd 2021-10-06 20:50:41 in which branch you have the build script? 2021-10-06 20:52:55 PJ[m]: Master 2021-10-06 20:54:02 yeah, I already got it 2021-10-06 20:55:07 Saijin_Naib[m]: you have builddir="$srcdir", which is wrong 2021-10-06 20:55:15 In this case, you can remove it 2021-10-06 20:55:19 because the default should work 2021-10-06 20:59:14 Oh, okay. Thanks! 2021-10-06 20:59:19 I think that is in the template 2021-10-06 21:00:13 and remove prefix before `Cargo.toml` 2021-10-06 21:00:33 Oh, is that wrong now since I dropped the builddir? 2021-10-06 21:00:36 Saijin_Naib[m]: the template should have builddir="$srcdir/$pkgname-$pkgver" 2021-10-06 21:01:09 Oh, then I have no idea how I messed that up, haha 2021-10-06 21:02:01 Oop, we got building 2021-10-06 21:02:07 Yooo, thank you both so much! 2021-10-06 21:02:17 ` Building [=========================> ] 82/83: lsd(bin)` 2021-10-06 21:02:23 will see how big bin is 2021-10-06 21:02:31 all my 4c work hard 2021-10-06 21:02:33 Wow, fast. You beat me to it already haha 2021-10-06 21:03:17 Hehe, I wonder how hard this would have beat up my N3450 2021-10-06 21:04:03 I have an AMD SoC 2021-10-06 21:04:08 4c 2GHz 2021-10-06 21:04:25 not bad! 2021-10-06 21:04:27 check /etc/abuild.conf JOBS to make sure it uses more cores 2021-10-06 21:04:36 although for rust it might not matter 2021-10-06 21:05:03 it will not respect `JOBS` 2021-10-06 21:05:15 It passed/finished build 2021-10-06 21:05:19 I would not have expected that 2021-10-06 21:05:20 But my directory is empty 2021-10-06 21:05:26 What directory 2021-10-06 21:05:33 The package ends up in ~/packages/ 2021-10-06 21:05:35 my ~/Sources/aports/testing/lsd one 2021-10-06 21:05:35 although I should have built it on my TR1920X 2021-10-06 21:06:05 which is also a valid repository you can add to /etc/apk/keys (~/packages/testing for example) 2021-10-06 21:06:06 Is this because I didn't change the behavior in the /etc/abuild.conf to not clear the directory after build? 2021-10-06 21:06:06 it should produce an apk in `~/packages` 2021-10-06 21:06:18 oooh 2021-10-06 21:06:34 Yep, there it is! 2021-10-06 21:07:08 So, install, and then stat the binary to get the size? 2021-10-06 21:07:28 can do 2021-10-06 21:08:12 File: /usr/bin/lsd 2021-10-06 21:08:12 Size: 1718320 2021-10-06 21:08:13 I have this which I don't remember how or when I installed it 2021-10-06 21:08:13 `.rwxr-xr-x root root 2.7 MB Wed Aug 25 00:30:28 2021  /usr/share/rust/.cargo/bin/lsd` 2021-10-06 21:08:36 By the manual cargo install method they have on their github maybe? 2021-10-06 21:08:42 So, this looks pretty significantly smaller 2021-10-06 21:08:52 Neat! 2021-10-06 21:09:19 It apparently can get smaller when Rust adds the strip symbols thing in stable 2021-10-06 21:10:59 So, at this point I should make a branch for lsd specifically, right? And then open the MR from that? 2021-10-06 21:11:19 correct 2021-10-06 21:11:39 Anyone want to help me struggle through GNOME Circle Drawing and Pinta? Haha 2021-10-06 21:11:45 well, you don't have to make a branch but it's a bad pattern to make pull/merge requests from master 2021-10-06 21:11:59 Saijin_Naib[m]: if you push it as a branch, I can help you 2021-10-06 21:12:00 Appreciate that. Makes sense, I guess 2021-10-06 21:12:26 PJ[m]: Sweet. I'll try to get those started soonish 2021-10-06 21:40:00 eh, setting up new alpine is painful... 2021-10-06 21:40:09 Lots of customizations? 2021-10-06 21:40:30 yes, such as `doas` which isn't that popular like `sudo` 2021-10-06 21:41:29 I need to automate/package some stuff, since I use simple go-to-all link `elevate` that points to whatever is installed 2021-10-06 21:41:42 and then patch `fish` so it doesn't prefix with `sudo` 2021-10-06 21:41:45 PJ[m]: does popularity mattwr? ;) 2021-10-06 21:41:48 hello 2021-10-06 21:41:53 it does a lot 2021-10-06 21:42:00 surely the concern is not popularity but utility 2021-10-06 21:42:24 doas is more useful on account of not being some eldritch monster summoned from the deep 2021-10-06 21:43:04 that still related to popularity 2021-10-06 21:43:25 reverse the roles and it would be the same for people who want more customisation 2021-10-06 21:43:56 ? 2021-10-06 21:44:29 "if `doas` was more popular than `sudo`" 2021-10-06 21:44:33 doas is less code and has equivalrnt functionality to sudo 2021-10-06 21:44:49 far less code 2021-10-06 21:44:59 the doas manual does not need a manual 2021-10-06 21:45:03 the sudo manual does 2021-10-06 21:45:20 Thanks, I was already aware of that 2021-10-06 21:45:30 this is a good property to have :) 2021-10-06 21:47:14 afaik, the functionality between both is not 1:1 2021-10-06 21:47:38 happy to be proven wrong though 2021-10-06 21:47:40 they do not have the same functionality 2021-10-06 21:47:58 and I don't mean just flags 2021-10-06 21:48:26 sorry to bother you, but has anything changed between alpine 3.14.2 and current edge that's related to having root partition encrypted with keyfile? 2021-10-06 21:48:28 Yeah, doas setup and making an alias bites me in the rear sometimes 2021-10-06 21:48:36 I can't get rid of the sudo muscle-memory 2021-10-06 21:49:16 ots not 1:1 no 2021-10-06 21:49:19 nor is it the same 2021-10-06 21:49:25 neither of which i claimed to 2021-10-06 21:49:42 doas is less code and has equivalrnt functionality to sudo 2021-10-06 21:49:45 > doas is less code and has equivalrnt functionality to sudo 2021-10-06 21:49:52 :> 2021-10-06 21:49:53 runit-12: should be fine, iirc there was an openssl issue but it was fixed 2021-10-06 21:49:53 equivalent != same... 2021-10-06 21:50:01 lol 2021-10-06 21:50:24 Having similar or identical effects. ;) 2021-10-06 21:50:42 neither have I said that they are same 2021-10-06 21:50:51 not claiming you are 2021-10-06 21:50:58 this language game is so boring 2021-10-06 21:51:04 I got it to work perfectly with 3.14.2, but as soon as I updated it stopped working, pretty sure I checked and everything about the required configuration is the same 2021-10-06 21:51:07 lets talk in binary 2021-10-06 21:52:18 anyway, I'll have to deal with that and make some packages so `sudo` and `elevate` nicely point to `doas` 2021-10-06 21:53:24 For me it is mostly just getting lsd, tlp, doas, and my DE setup 2021-10-06 21:53:40 If I could get iwd to play nice all the time, I'd like to automate that for setup as well 2021-10-06 21:54:31 huh, i just use wpa_supplicant 2021-10-06 21:55:23 never used `iwd`, `wpa_supplicant` *almost* always worked for me well 2021-10-06 21:57:05 would be nice if `pihole` guys give a bit more care to Alpine 2021-10-06 22:09:16 o/ 2021-10-06 22:13:36 o/ 2021-10-06 22:23:48 "huh, i just use wpa_supplicant" <- Supposedly iwd is meant to replace wpa_supplicant, but it has been an absolute nightmare for me 2021-10-06 22:24:03 Do pihole not contribute much upstream? How is OpenWRT in that regard? 2021-10-06 22:47:27 honest question: Other than a php web front-end, what would you like to be added to https://wiki.alpinelinux.org/wiki/Using_Unbound_as_an_Ad-blocker 2021-10-06 22:47:40 to make it a "pihole for cheap"? 2021-10-06 23:10:05 think people just like seeing those fancy statistics on the frontend 2021-10-06 23:11:16 hi 2021-10-06 23:12:00 is there a way to use older linux kernel but still user new packages from apt-get ? 2021-10-06 23:12:09 alpine doesn't have an apt-get 2021-10-06 23:12:23 whatever apt analog it has 2021-10-06 23:12:45 not unless you compile the kernel yourself 2021-10-07 00:39:27 hi everyone, i decided to do somethign crazy today, but i dont know where to start and could use a pointer or two. (im doing this for fun, not for production, so no need to worry about security)... how can i modify a stock alpine install to boot straight to shell without any login or password 2021-10-07 00:39:41 eventuall maybe even boot straght to an application screen 2021-10-07 00:39:53 edit inittab 2021-10-07 00:40:04 i dont need step by step help, but some bullet points will help if anyone is kind enough 2021-10-07 00:40:23 what items should i look for in inittab 2021-10-07 00:50:41 ok well that helped thanks i figured it out!! appreciate it 2021-10-07 01:24:41 Is there an official bittorrent file or magnet for getting the ISOs? 2021-10-07 01:35:56 i don't think there is an official one as it's not mentioned anywhere near official downloads 2021-10-07 01:36:06 but the sha256 sums are published and should match anyway 2021-10-07 01:36:16 ditto for a gpg signature 2021-10-07 01:36:51 if it's just your preference in downloading via bittorrent and verifying later, that is 2021-10-07 07:08:11 Hello, I heard that there are some global issues with expired wolfssl certificate, not sure this affects alpine, but being based on busybox too, I assume so. Since today I'm getting such errors as "wget: can't execute 'ssl_client': No such file or directory" curl also doesn't work. Is there a workaround currently for these errors? 2021-10-07 07:11:37 correction, it's letsencrypt's global root cert which is expired and causing quite a stir also over at the openwrt community 2021-10-07 08:45:42 hi all 2021-10-07 08:45:57 I have a question about Alpine in diskless mode 2021-10-07 08:46:07 I can't find this info anywhere 2021-10-07 08:46:38 I hava an USB key with alpine set up ad diskless with the script setup-bootable 2021-10-07 08:46:47 and works perfectly 2021-10-07 08:47:08 now I need to customize the kernel-line parameters 2021-10-07 08:47:12 there is a way ? 2021-10-07 08:47:23 or I have to accept the ISO defaults ? 2021-10-07 08:47:47 I know there is a script update-kernel 2021-10-07 08:48:04 but I can't find any pointers to the kernel-line parameters 2021-10-07 08:49:36 usually, in "sys" mode, those are in /etc/update-extlinux.conf (if booted legacy) or in the grub config 2021-10-07 09:33:32 I've noticed that when you run the update-kernel script, in the usb key the /boot dir is created 2021-10-07 09:33:48 maybe it's an overlay for the boot process? 2021-10-07 09:34:05 can't find nay docs anywhere 2021-10-07 09:59:03 wyk72 - on my usb key theres a file in / called syslinux.cfg; sometimes it can be somewhere under boot as well. the line you are looking for is APPEND 2021-10-07 09:59:16 that's where you add kernel command-line parameters 2021-10-07 10:00:11 IIRC you may have an extlinux.cfg file instead 2021-10-07 10:11:03 I solved by running the upgrade-kernel script and then editing manually the /media/usb/boot/grub/grub.cfg there (or syslinux) 2021-10-07 11:54:40 Hello! I'm trying to install Alpine linux behind proxy, but wget is falling to get apk repositories. Any ideas ? Thanks 2021-10-07 11:56:13 even with http_proxy and https_proxy set? 2021-10-07 11:59:14 Yes, even with http_proxy and https_proxy set in env 2021-10-07 12:12:28 failing how? 2021-10-07 12:18:37 setup-apkrepos - wget: error getting response: Invalid argument wget: error getting response: Resource temporarily unavailable 2021-10-07 12:19:27 And a normal wget from the commandline (with *_proxy set) works? 2021-10-07 12:20:50 Nope, but same setup with other linux box is working, proxy is working 2021-10-07 12:24:28 Never mind, I'll try with direct connection. Thank you all. 2021-10-07 12:28:02 uppercasing those env vars might help 2021-10-07 13:05:14 hi i have some coredump I want to inspect but apparently the instruction pointer is at musl, so no debug symbols. I read there's musl-dbg, but how do I use that? isn't my program still linked to /lib/ld-musl-x86_64.so.1 2021-10-07 13:11:38 agsv: just install it. They are detached debug symbols 2021-10-07 13:57:43 ikke: can I use them on dumps that came from a system that didn't have that package installed? 2021-10-07 14:00:55 as long as the system had the same musl/musl-dbg version and arch i think 2021-10-07 14:01:33 in docker we trust 2021-10-07 14:05:32 Yes, you can install the symbols after the fact 2021-10-07 14:24:26 semi-related i can't seem to make gdb to load musl into the debugging session, it might be some lack of knowledge of gdb on my side, but i'm asking here in case it's some alpine quirk you guys might help me with 2021-10-07 14:24:30 http://ix.io/3B6q 2021-10-07 14:25:40 the internet claims I provided bad coredump, but I have some other sample that crashes in openresty code, and in that case gdb manages to figure the right function that crashed. 2021-10-07 15:38:59 does anyone here use gdm on Alpine? I get an error like this: https://gitlab.alpinelinux.org/alpine/aports/-/issues/11347#note_172631 2021-10-07 15:39:16 except vt 1 instead of 7 2021-10-07 15:40:57 I'm experiemting that problem Newbyte 2021-10-07 15:41:22 I tried to fix it but didn't get any success 2021-10-07 15:41:24 donoban: do you know any workaround? 2021-10-07 15:41:31 I see. thanks for trying 2021-10-07 15:43:55 I have some log with gdm debug enabled but I think that there is no useful info 2021-10-07 15:44:14 to me it sounds more like an X.Org configuration issue 2021-10-07 15:45:59 I'm pretty sure that it broken with some unrelated package upgrade 2021-10-07 15:46:06 (unrelated to gdm/pam) 2021-10-07 15:46:38 probably 2021-10-07 15:46:50 maybe xorg-server? 2021-10-07 15:47:00 community/xorg-server: upgrade to 1.20.13 2021-10-07 15:47:10 I wonder if it would help to make it run via Wayland 2021-10-07 15:47:40 could I downgrade xorg-server package? 2021-10-07 15:52:10 donoban: are you using gdm? 2021-10-07 15:52:33 no but I tried to improve it's pam configuration 2021-10-07 15:52:58 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/25178 2021-10-07 15:53:11 when I created that MR it worked 2021-10-07 15:53:36 donoban: did it work without the config changes too? 2021-10-07 15:54:04 well, the reallity is that now it doesn't work with old/new config 2021-10-07 15:54:10 so I think that the pam config has no relation 2021-10-07 15:54:12 oh, I see 2021-10-07 15:54:38 my guess is that it's some issue with rootless xorg 2021-10-07 15:54:59 I'm trying to rebuild xorg-server with older version 2021-10-07 15:55:11 did you try making the xorg executable setuid? 2021-10-07 15:55:39 uhM no 2021-10-07 15:56:01 let's try 2021-10-07 15:56:06 Xorg? 2021-10-07 15:57:11 something like that 2021-10-07 15:57:46 rebooting.. 2021-10-07 15:58:51 why? 2021-10-07 15:59:31 because I already had an sway session opened there 2021-10-07 15:59:36 fastest way for test 2021-10-07 15:59:54 ah, all right 2021-10-07 16:00:01 "does anyone here use gdm on..." <- Was using gdm until about a month ago, when it just went to black screen with white text cursor in top left no matter what I did. Nuked my GNOME+GDM install for XFCE4+LightDM 2021-10-07 16:00:15 Yeah, that's what I get too 2021-10-07 16:00:27 I wouldn't recommend using Matrix replies here btw, they look ugly on the IRC side I've herad 2021-10-07 16:00:29 heard* 2021-10-07 16:00:29 it stills failling 2021-10-07 16:00:33 but maybe I didn't set setuid proplery 2021-10-07 16:01:26 Ugh... I mean... Why have the client/protocol then if I'm not supposed to use it? 2021-10-07 16:01:31 Like... I don't get it 2021-10-07 16:01:38 This is an IRC room 2021-10-07 16:02:06 Can the Matrix bridge not be configured to strip Matrix-only stuff? 2021-10-07 16:02:12 apparently not 2021-10-07 16:02:16 Sigh 2021-10-07 16:02:17 I mean, IRC users can read replies just fine 2021-10-07 16:02:21 Alright, I'll try to be mindful 2021-10-07 16:02:22 It just doesn't look pretty 2021-10-07 16:02:42 It's not like your message doesn't get send or anything 2021-10-07 16:02:43 -rwsr-sr-x does this means suid? 2021-10-07 16:03:03 What happens with code-blocks or other formatting? Also a mess? 2021-10-07 16:03:15 It sends a link to the multiline message 2021-10-07 16:03:16 They get a link to the message if it's multiline 2021-10-07 16:03:22 Regular formatting like bold and italic just get send as is 2021-10-07 16:03:34 Some people complain about that too for some reason (I don't get why, I can't think of a better way to handle it) 2021-10-07 16:03:53 Oh, was that the whole don't send more than X characters thing the other day? 2021-10-07 16:03:56 Me neither, alternative is a pastebin like service which is just the same but with more work on the sender part 2021-10-07 16:04:05 Not sure I have a client that gives me column count 🙄 2021-10-07 16:04:12 Saijin_Naib: I guess so, but it does it on multiline not on a set amount of characters 2021-10-07 16:04:16 Starting to sound like mailing list "etiquette" 2021-10-07 16:04:17 Ahh, okay 2021-10-07 16:04:32 IRC doesn't support multiline so alternatives methods are used 2021-10-07 16:05:04 One would think that the bridge would just split on newline and send as multiple messages concurrently 2021-10-07 16:05:18 Maybe that's a harder problem to solve than it seems, like everything else haha 2021-10-07 16:06:14 Hmm... Maybe it is Matrix/Element's onus, then, if they know you're using a bridge, to "limit" the client to IRC-only behaviors 2021-10-07 16:06:27 Newbyte: interesting, now instead a black screen I get a nice graphic error "Wops, something went wrong..." (maybe this is pam related) 2021-10-07 16:06:42 but I need to suid /usr/libexec/Xorg 2021-10-07 16:06:48 donoban: wow, that's progress! 2021-10-07 16:06:51 Did you add the gdm user to the tty group, or your login user? 2021-10-07 16:07:21 I have my gdm as part of audio,video,plugdev already 2021-10-07 16:07:21 Saijin_Naib: that would get abused quickly and triggers spam filters. You can just fill the chat with a long enough message 2021-10-07 16:07:21 I tried but removed after don't getting any help 2021-10-07 16:07:51 I don't think that you need to add it to non-default groups 2021-10-07 16:07:53 PureTryOut (matrix.org): Of course. 2021-10-07 16:08:08 By default I think gdm was only a member of its own group 2021-10-07 16:09:37 it should be ok, it worked fine before this problem 2021-10-07 16:11:19 ouch there is too much gdm log :( 2021-10-07 16:11:52 at least Xorg is alive 2021-10-07 16:16:05 "WARNING: App 'org.gnome.Shell.desktop' exited with code 1" maybe this 2021-10-07 16:16:38 Newbyte: have you tried with xorg suid? 2021-10-07 16:20:20 donoban: no, not yet 2021-10-07 16:28:27 I'm gonna try playing with permissions 2021-10-07 16:31:44 black screen again :\ 2021-10-07 16:46:02 @donoban what did you change? 2021-10-07 16:46:21 removed the suid, added gdm to tty group, changed all tty* to root:tty 2021-10-07 16:58:23 hello, I remember some time ago there was a discussion about removing firejail from alpine, is anyone willing to review my (poor) attempt at making a firejail clone ? 2021-10-07 17:02:30 why not use/contribute to bubblewrap instead? 2021-10-07 17:03:08 uhM, there is a firejail clone based on bubblwrap, bubblejail 2021-10-07 17:03:11 bubblewrap uses command line parameters 2021-10-07 17:03:23 I don't like bubblejail syntax 2021-10-07 17:03:45 well, your clone works without need of suid? 2021-10-07 17:03:53 yup 2021-10-07 17:04:26 though I haven't really added a lot of features so I don't know if you can go all the way without suid 2021-10-07 17:05:36 also bubble jail is in python, I would prefer not to install python just for sandboxing 2021-10-07 17:06:58 yeah sounds reasonable for writing a clone, just asked for suid because in case of need it is not very likely that it would be accepted 2021-10-07 17:08:02 I have to be hones I don't understand why firejail needs suid, or any recent-ish linux project 2021-10-07 17:09:36 maybe it can be "fixed" but it could require a lot of work 2021-10-07 17:10:54 yeah, large projects are very hard to properly maintain 2021-10-07 17:20:26 tiotags: what mechanism does your "clone" use for sandboxing? 2021-10-07 17:21:01 namespaces ? 2021-10-07 17:22:33 I want to implement seccomp filters and others later 2021-10-07 17:24:06 only mount namespaces atm, it's really early, atm it's just a text parser and 'mount things in a virtual folder' 2021-10-07 17:26:53 Newbyte if you're interested in how a script looks like https://gitlab.com/tiotags/hardshell/-/blob/master/work/test.sh 2021-10-07 17:41:49 donoban: how did you setuid it to get the oh no screen? 2021-10-07 17:42:29 chmod u+s /usr/libexec/Xorg 2021-10-07 17:42:51 and also /usr/bin/Xorg but I don't know if it was executed 2021-10-07 18:39:18 I wonder if we could work around this by just running gdm in Wayland mode 2021-10-07 18:39:33 It shouldn't depend on rootless X.Org working 2021-10-07 18:51:41 I'm checking it 2021-10-07 18:51:55 Use WaylandSupport : YES 2021-10-07 18:52:35 it's the current config, but it seems that it doesn't use it 2021-10-07 18:56:29 Weird, I wonder why 2021-10-07 19:06:48 I don't see any reference to wayland on the log 2021-10-07 19:08:19 You probably need to configure it somehow, I know you need to with the new Wayland support in SDDM for example 2021-10-07 19:09:58 https://wiki.archlinux.org/title/GDM#GDM_ignores_Wayland_and_uses_X.Org_by_default 2021-10-07 19:10:40 it doesn't seem our problem 2021-10-07 19:22:02 uhM, I noticed that each gdm start attemp causes: 2021-10-07 19:22:09 [ 1407.181149] nouveau 0000:01:00.0: Enabling HDA controller 2021-10-07 19:22:11 [ 1407.327543] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 6013d4 [ IBUS ] 2021-10-07 19:25:23 PureTryOut: I don't see any option except WaylandEnable=true 2021-10-07 19:26:01 🤷 2021-10-07 19:39:21 "gdm already has Conflicts/After on getty@tty1.service" 2021-10-07 19:39:55 uhM, could be this? 2021-10-07 19:42:48 beh, nothing 2021-10-07 20:14:06 h non 2021-10-07 20:14:11 wrong buffer 2021-10-08 13:22:17 What is the bare minimum set of files needed in a folder in order for apk to be able to add packages? I know doing a lbu package into the folder works but I'm wondering which files are absolutely essential for apk to work. 2021-10-08 13:22:54 a local apk repo? 2021-10-08 13:24:09 lopid: Assuming the system running the apk add has internet access is that needed? 2021-10-08 13:24:33 There is apk-static 2021-10-08 13:25:08 sounds like XY problem 2021-10-08 13:32:12 theoretically apk should have a complete set of dependencies 2021-10-08 13:32:21 including the abuild auto-generated deps 2021-10-08 13:33:23 so something like apk info -W $(apk info -R apk-tools), but recursive 2021-10-08 13:33:28 s/-W/-L/ 2021-10-08 13:34:28 So looking at the alpine-chroot-install script, what's needed is the static apk-tools, apk repositories and apk keys? 2021-10-08 13:34:50 https://github.com/alpinelinux/alpine-chroot-install/blob/3ceb85144247b7b193f8bc27c21c3d8895427088/alpine-chroot-install#L322-L345 2021-10-08 13:35:50 sounds plausible 2021-10-08 13:38:14 Is apk-static something special or is it ok to use the normal apk ? (The alpine-chroot-install script doesn't seem to have any different static apk) 2021-10-08 13:38:39 it's compiled statically in contrary to standard apk 2021-10-08 13:39:20 ktprograms: the main purpose/benefit of apk-static is that it can be run on a non-Alpine system to install in a chroot 2021-10-08 13:50:30 So to make an alpine chroot on alpine I can use the apk from the host? 2021-10-08 13:51:16 Yes 2021-10-08 13:53:24 ktprograms: for example I can build Alpine disk images on a Debian machine using apk-static 2021-10-08 14:10:57 to make alpine chroot from alpine you need only do apk add --root alpine alpine-base 2021-10-08 14:11:19 possibly also need -X, but it should use host keys by default 2021-10-08 14:14:55 I get 'Unable to open lock database: No such file or directory'. (The only files I have in the chroot dir is etc/apk/repositories and etc/resolv.conf) 2021-10-08 14:16:10 If I use --initdb it doesn't have that error but it says UNTRUSTED Signature 2021-10-08 14:20:08 In the end this is the command I needed in order to get a working chroot: apk add -p chroot --initdb --allow-untrusted alpine-base 2021-10-08 14:21:15 I typically use "apk --arch x86_64 --initdb --allow-untrusted --root chroot-dir --update-cache add alpine-base" 2021-10-08 14:22:11 plus adding some other packages such as grub/syslinux as I'm creating disk images 2021-10-08 14:23:46 you can avoid the need for "--allow-untrusted" if you copy the Alpine keys into etc/apk/keys inside the chroot. 2021-10-08 14:26:44 minimal: Thanks that works, but any idea why I have 2 .pub files in my /etc/apk/keys folder? 2021-10-08 14:35:36 ktprograms: not sure - I guess it depends on the Alpine version you've installed in the chroot (some packages signed by different key than others?) 2021-10-08 14:35:57 are they both "alpine-devel" keys? 2021-10-08 16:51:27 Saijin_Naib[m]: I saw your package come through on pkgs.alpinelinux.org. Congrats on submitting your first package! 2021-10-08 17:03:10 xordspar0 thanks! Definitely caused some folks some heartburn 😬 2021-10-08 17:03:17 Hopefully a few more soon 2021-10-08 17:04:26 hopefully no more heartburns :P 2021-10-08 17:12:22 Yeah, that's an ideal 2021-10-08 17:44:23 how to assign multiple IPv6 addresses to one interface? 2021-10-08 17:45:20 you can repeat the address clause multiple times if you use ifupdown-ng 2021-10-08 17:52:22 you have served your purpose 2021-10-08 18:20:24 Are there packages somewhere/ports for LLVM 12/13? 2021-10-08 18:20:37 aports seems to only go up to 11 2021-10-08 18:46:56 da_lorax: Spinal Tap moment? ;-) 2021-10-08 18:47:18 lol 2021-10-08 18:50:42 ikke: could have made a Burnistoun reference but not many people would get that lol 2021-10-08 18:50:58 I at least would't have 2021-10-08 18:51:23 ikke: Scottish accent joke, https://www.youtube.com/watch?v=sAz_UvnUeuU 2021-10-08 18:51:36 ah, I know that one :) 2021-10-08 18:51:53 Just did not know the name 2021-10-08 18:52:29 "it don't do scottisch accents" 2021-10-08 18:54:03 freedom! 2021-10-08 18:54:39 :D 2021-10-08 18:55:32 one of those accents people find hard to do, like Irish accents, and that usual in films even native-speaking actors put on a "Hollywood" version of the accent 2021-10-08 19:04:38 da_lorax im working on 12 atm 2021-10-08 19:04:52 Just gotta fix some packages that broke 2021-10-08 19:05:20 But i do not have time this weekend 2021-10-08 19:27:54 Misthios: need help? I've got time this weekend :) 2021-10-08 19:34:59 this works with ed from repo, but not from busybox, isnt it suppesed to ? https://pastebin.com/840Hk7SJ 2021-10-08 19:35:22 i get no errors 2021-10-08 19:36:07 actually, when i run it with 'sh' (isntead of zsh), i get : ""/1/w/ib", Illegal instruction" 2021-10-08 19:36:40 da_lorax if u want u can look at !25551 2021-10-08 19:37:58 My statusupdate is out of date, see job logs for up to date version (s390x has failing postgre due to jit, and all of the gbvm or whatever fail due to the ssl1.1-compact vs ssl3 conflict) 2021-10-08 19:56:02 so many versions of the same program creates problems eh 2021-10-08 19:57:18 Well, incosistent dependencies cause problems 2021-10-08 20:24:23 does anyone here use gdm on Alpine? 2021-10-08 21:42:21 hi, I noticed linux-edge builds the 'wacom_w8001' module but does not build the 'wacom' module. The latter is much more useful and common, because the former only supports older devices (like >10 years old), while the latter is needed for anything newer. Is there a reason this is so, and if not, what can I do to change this? 2021-10-08 21:44:19 Is that CONFIG_HID_WACOM? 2021-10-08 21:49:46 yes 2021-10-08 21:50:20 Right, that's not set at the moment 2021-10-08 21:50:28 You could create a feature request for it 2021-10-08 21:51:06 how does one do that? 2021-10-08 21:52:50 https://gitlab.alpinelinux.org/alpine/aports/-/issues 2021-10-08 21:52:53 There you can create an issue 2021-10-08 21:53:16 thanks 2021-10-08 22:22:36 oksica 2021-10-08 22:22:39 oops 2021-10-09 00:28:44 minimal: They're both alpine-devel keys. 2021-10-09 00:58:25 Along those lines, what are the general ideas behind enabling devices for linux-edge vs linux-lts? For instance, I got a MR approved for enabling the i2c over HID module and the pinctrl_broxton module for supporting the touchpad on intel reference laptops built around the Celeron quad-core CPUs for linux-lts. Would that not be MR material for linux-edge? 2021-10-09 00:58:47 Is edge less about hardware enablement/support and more about just testing new kernel changes as they come from upstream? 2021-10-09 01:07:13 It turns out that in order to make a different architecture chroot, you need to have the qemu-$arch binary in the chroot folder before the apk add alpine-base 2021-10-09 01:07:34 Wonder if I should write a wiki page on making a cross architecture chroot? 2021-10-09 01:09:56 I seem to get 'ip: OVERRUN: Message too large' when running 'ip a' in the chroot. I've mounted proc, sys, dev and dev/pts 2021-10-09 02:11:48 More documentation is never a bad thing 2021-10-09 02:24:22 please put it anywhere other than the wiki, the wiki is a graveyard of obsolete and outright incorrect material 2021-10-09 02:24:41 even a blog is better, then people can easily see who wrote it and when 2021-10-09 02:25:22 for cross-arch you should set binfmt_misc flags PO. i think qemu-binfmt is supposed to do this already though 2021-10-09 02:32:29 Hello71: Is there a qemu-binfmt package on alpine? 2021-10-09 02:33:14 dunno if it's a separate package or what it's called exactly but apk add qemu-openrc; /etc/init.d/binfmt start 2021-10-09 02:33:16 or something like that 2021-10-09 02:33:29 i think it is in qemu-openrc 2021-10-09 02:33:41 oh, no, it is /etc/init.d/qemu-binfmt 2021-10-09 02:34:14 ah, P is not needed, it is F 2021-10-09 02:34:50 which implies O 2021-10-09 02:35:16 regarding ip, if you run any netlink programs under qemu don't be surprised if they don't work 2021-10-09 02:35:47 if your networking doesn't work then you've failed to set up /etc/resolv.conf, that's the only thing that needs to be done in chroot 2021-10-09 02:36:02 Hello71: Networking works (can apk add) 2021-10-09 02:36:09 chroot keeps same networking, that's why it is called ch*root* 2021-10-09 02:36:45 theoretically ip a should work but the netlink protocol is very complicated and i don't think qemu even tries to implement it 100% 2021-10-09 02:37:15 Hello71: I see 2021-10-09 02:37:55 i can't imagine it being used for anything other than ifupdown-ng which won't work right in chroot anyways 2021-10-09 02:38:13 and crappy proprietary licensing software which is probably also broken in chroot 2021-10-09 06:27:45 Is there some kind of way to define two versions of a package at once? I'm tryna get this llvm12 MR working and https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/25551/diffs#df9529dfe32498ea305cb8cff23ffe44eaaa950e_10_10 looks odd 2021-10-09 06:28:15 I'm new to alpine so unsure if magic I don't know or a mistake 2021-10-09 06:45:22 Do u mean the arch= all !x86 line? 2021-10-09 06:45:47 If so that just tells apk to build each arch except the ones marked with the ! 2021-10-09 07:08:59 Misthios: no the fact that all the variables are there twice 2021-10-09 07:09:09 in lldb 2021-10-09 07:09:19 like, it adds the new versions but doesn't look like it deletes the old one 2021-10-09 07:09:39 also where is the "job log" that shows what else needs done-ing? 2021-10-09 07:14:57 i.e. what needs to pass in order for it to be finished? 2021-10-09 09:06:04 There is a pipelines tab 2021-10-09 09:06:42 https://gitlab.alpinelinux.org/Misthios/aports/-/jobs/507114 2021-10-09 09:09:27 In the end there is a summary of failed packages 2021-10-09 11:07:58 Hello. I use AlpineWSL 3.14.0 and cannot install a pip for python2. 2021-10-09 11:07:58 apk add py-pip install me python3 like py3-pip.. and finaly I cannot pip install requests for pytho2 because I have only python3 and I stop in dead point 2021-10-09 11:09:23 SiB: python2 has very limited support left in alpine, mostly because it's required as a build dependency for some larger packages 2021-10-09 11:09:36 So all non-essential things are no longer provided 2021-10-09 11:10:25 ikke: Maybe some workaround... older Apline if exist for WSL ? 2021-10-09 11:14:42 I found that Alpine 3.5-3.11 have py2-pip but this is possible to back? probably not bcs repo do update to never package and I am in loop, true ? 2021-10-09 11:14:57 `python -m ensurepip --upgrade` shouldn't that work if you have py2? 2021-10-09 11:16:12 PJ[m]: Successfully installed pip-19.2.3 setuptools-41.2.0 | Thanks, Now I go check if this solve main problem ! 2021-10-09 11:56:00 PJ[m]: This solve all my problems, THANKS 2021-10-09 11:57:24 👍 2021-10-09 12:36:56 except your security vulnerabilities... 2021-10-09 15:18:36 is this the only date/time format maildir accepts? "Mon, 27 Sep 2021 09:55:06 +0200" (was surprisingly hard to search for) 2021-10-09 15:40:38 RFC somesuch 2021-10-09 17:52:10 minimal: Did your updates to create-alpine-disk-image bring it to where you want? 2021-10-09 18:04:22 talismanick: Hi. Its still a work in progress. I keep meaning to push my local updates to the GitHub repo. 2021-10-09 18:04:49 as you've reminded me, let me do that now :-) 2021-10-09 18:07:43 talismanick: was there some particular aspect of it that you are interested in? 2021-10-09 18:08:40 minimal: Oh, I was going to use it to make myself a VPN-on-a-VPS, but you said you'd push new changes soon 2021-10-09 18:08:44 so I waited 2021-10-09 18:09:44 I've been busy, anyways - I just now have time to take another shot at doing it right 2021-10-09 18:09:48 sorry, too much procrastination ("let me add one more thing before pushing...") :-) 2021-10-09 18:10:09 minimal: I suppose you could say that's been my approach to setting up my VPN 2021-10-09 18:10:16 just updating the CHANGELOG now before doing a push 2021-10-09 18:36:06 talismanick: ok, changes merged 2021-10-09 18:36:20 minimal: awesome :D 2021-10-09 18:36:29 thanks for all the work 2021-10-09 19:29:51 is there anywhere alpine logo without text with good resolution (or svg)? 2021-10-09 19:38:35 JuniorJPDJ: This is the logo as SVG: view-source:https://alpinelinux.org/alpinelinux-logo.svg 2021-10-09 19:38:47 It includes the text, but should be easy enough to remove 2021-10-09 19:41:40 Wiki seems to have nice one without text :D 2021-10-09 19:41:43 thanks anyway! :D 2021-10-09 19:42:01 good 2021-10-09 19:42:35 It's not svg, though 2021-10-09 19:59:56 good enough for my purposes ;) 2021-10-09 20:00:07 just wanted to have logo on matrix space 2021-10-10 04:05:38 What in the world 2021-10-10 04:06:51 Hello, does someone here use Anbox on Alpin 2021-10-10 04:07:16 s/Alpin/Alpine? I have an ARM64 machine and want to get it working on ARM64./ 2021-10-10 06:37:03 Hi, I've made a How-To on the Alpine Wiki, but I get this message when I click save: 'Error: This action has been automatically identified as harmful, and therefore disallowed. If you believe your action was constructive, please inform an administrator of what you were trying to do. A brief description of the abuse rule which your action matched is: Spammers not welcome'. My wiki username is 2021-10-10 06:37:09 'Ktprograms' if that helps. Thanks! 2021-10-10 06:44:10 I just realised this question should have been in #alpine-docs. Sorry! 2021-10-10 07:36:22 /b qut 2021-10-10 13:03:48 errweas[m]: i vaguely recall that some people use it on pmos, but alpine is missing some kernel config options. what issues are you having exactly 2021-10-10 13:39:22 I cannot load the needed binder_linux module since Alpine doesn't have dkms 2021-10-10 13:39:42 Ashmem is active though 2021-10-10 13:39:49 errweas[m]: we have akms nowadays 2021-10-10 13:39:54 oh 2021-10-10 13:39:55 ok 2021-10-10 13:39:59 https://github.com/jirutka/akms 2021-10-10 13:40:21 good, I will proceed 2021-10-10 13:46:26 ACTION uploaded an image: (4915KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/xIQYgoNFBeNjIUjKQFIgqJFA/20211010_154559_6498175741973537278.jpg > 2021-10-10 13:47:30 Why root?: My xorg freezes immediately after it's entered on user account. 2021-10-10 13:52:39 you can't just replace dkms with akms and have it work 2021-10-10 13:52:42 the errors tell you that 2021-10-10 13:52:53 I did replace it. 2021-10-10 13:53:31 I put akms instead of dkms in the INSTALL.sh script. 2021-10-10 13:53:40 that does not work 2021-10-10 13:53:45 you can't replace `dkms install` with `akms install`, i know you did 2021-10-10 13:53:56 you need to make an AKMBUILD file for each module 2021-10-10 13:54:46 Oh oj 2021-10-10 13:54:49 lol 2021-10-10 13:54:54 * Oh ok. 2021-10-10 13:55:04 https://git.alpinelinux.org/aports/tree/testing/rtl8821ce-src/APKBUILD 2021-10-10 13:58:24 errweas: btw I recommend Waydroid rather than Anbox. Doesn't help with the kernel config problem you're having (as Waydroid has that too) but it'll perform way better than Anbox 2021-10-10 13:59:13 So that is also available on ARM64? 2021-10-10 14:06:48 Yes 2021-10-10 14:07:02 I see, you're the maintainer of waydroid, will Waydroid install an arm64-image? 2021-10-10 14:07:53 Currently the package will pull in waydroid-image which installs an arm64 image. I'll soon get rid of the waydroid-image package in Alpine though and then waydroid itself will pull in the image at runtime 2021-10-10 14:08:17 Awesome 2021-10-10 14:08:53 Does one have to be non-root to execute waydroid? 2021-10-10 14:09:05 It has part root part non-root components 2021-10-10 14:09:27 My user account xorg isn't working as of now 2021-10-10 14:09:31 You need to start the container, for which you can use the OpenRC service and requires root, and the session which is the non-root part 2021-10-10 14:09:34 But Waydroid requires Wayland 2021-10-10 14:09:57 I see I am not on Wayland 2021-10-10 14:10:41 Then you'll be stuck to Anbox. Or move to Wayland, it's the future after all 😜 2021-10-10 14:12:19 I use a special window manager setup with terminal blurring and opacity "dwm-flexipatch" I know there is dwl, (dwm for wayland) but I want my setup. 2021-10-10 14:12:49 ACTION uploaded an image: (3425KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/SJHqfcOrrQcukGRfILWpAHOW/IMG_20210913_215201.jpg > 2021-10-10 14:12:59 I use Gentoo regularly, btw. 2021-10-10 14:13:47 errweas[m]: You know actual screenshots are a thing? 2021-10-10 14:14:00 Yes 2021-10-10 14:14:19 My picture says enough. 2021-10-10 14:14:37 not sure what it's supposed to be saying 2021-10-10 14:15:09 That they use Gentoo :P 2021-10-10 14:15:26 distros are horoscopes for linux users 2021-10-10 14:15:56 what is cancer then? 2021-10-10 14:16:00 I suppose it's the rgb they wanted to show off 2021-10-10 14:16:09 They? 2021-10-10 14:16:24 They as in you 2021-10-10 14:16:36 Forget about the rgb, I didn't have a screenshot lurking around on my phone. 2021-10-10 14:16:36 They as in "I don't know your gender but people on the internet tend to get offended if I assume the wrong gender" 2021-10-10 14:16:44 PJ[m]: I see. 2021-10-10 14:16:57 PureTryOut: Ooh, ok 2021-10-10 14:16:59 I find it weird too but hey 🤷 2021-10-10 14:17:11 well, it's a proper use of English if you use "they" 2021-10-10 14:17:29 Anyway, I wanna get wayland working for Alpine. 2021-10-10 14:17:42 me too, buddy, me too 2021-10-10 14:17:43 PJ: doesn't make it less weird for me 2021-10-10 14:17:55 In my own language it's the local equivalent for they as well, still weird 2021-10-10 14:18:22 Because it's used for multiple people since the word existed, and now it can suddenly be used for a single person as well 🤷 2021-10-10 14:18:37 if by suddenly you mean a few hundred years 2021-10-10 14:18:50 We actually have singular version of "they" in my language 2021-10-10 14:20:10 The only issue is that 'they' can be ambiguous 2021-10-10 14:20:10 ok maybe not that sudden, but it's definitely way more used this way in recent years. A few hundred years ago people wanting to be called like that were still burned at the stake, so I doubt the use was used like that back then. And that's ignoring the fact that languages evolve so the word probably didn't exist in the current form anyway back then 😜 2021-10-10 14:20:18 Exactly that ^ 2021-10-10 14:20:44 Anyway 2021-10-10 14:21:04 Wayland works fine on Alpine, it just depends on your window manager and/or DE 2021-10-10 14:21:40 My picture was a dick-ish move, sorry. 2021-10-10 14:21:46 Was it? 2021-10-10 14:21:50 nothing dickish about it 2021-10-10 14:21:58 In my opinion 2021-10-10 14:22:09 It's not weird for me nor I won't be angry if someone asks me to use that word, but I'm also not behind idea of crusading with pitchforks for any mis-pronoun-ciation 2021-10-10 14:22:12 Well, that's just like... your opinion man... 2021-10-10 14:22:47 (responding to errweas) 2021-10-10 14:23:00 imfo, your picture was just irrelevant but cool setup you got there ;) 2021-10-10 14:23:41 PureTryOut: Right, but anyway I admire your work on Github especially the Anbox and waydroid stuff. I wanted to get Android emulation as I could ditch my phone. I want to do banking with android emulation. 2021-10-10 14:23:52 PJ[m]: Thanks, buddy. 2021-10-10 14:23:58 * Thanks, buddy :). 2021-10-10 14:24:41 Over a thousand hours of fixing issues and making a dream- system out of it, went into it :). 2021-10-10 14:24:58 My work on Github? I don't do much work on Github, what in particular are you talking about? 2021-10-10 14:25:08 Waydroid 2021-10-10 14:25:14 slaps Windows on PC and deals with it 2021-10-10 14:25:27 What work on Waydroid did I do on Github...? 2021-10-10 14:25:37 PJ[m]: That's not how the game is played. 2021-10-10 14:26:21 PureTryOut: You are the maintainer of waydroid. Or did you just fork it? 2021-10-10 14:26:38 Oh in Alpine. That is not done on Github lol, Github is just a mirror for aports 2021-10-10 14:26:52 Actual work is being done on https://gitlab.alpinelinux.org/alpine/aports 2021-10-10 14:27:26 PureTryOut: I see. 2021-10-10 14:32:59 How can I start wayland? 2021-10-10 14:33:31 `sway`? 2021-10-10 14:33:38 * `sway` ? 2021-10-10 14:38:01 errweas[m]: just type sway from a VT 2021-10-10 14:38:14 it does need a dbus session so its: dbus-run-session sway $@ 2021-10-10 14:38:36 the correct way would be to make a do_ script that exports all the variables needed 2021-10-10 14:40:34 ACTION uploaded an image: (5310KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/rmfWaCilAISDTtJAeuvhdtsL/IMG_20211010_164010.jpg > 2021-10-10 14:41:07 you don't need the $@ 2021-10-10 14:41:23 you are missing mesa-egl 2021-10-10 14:41:26 Ok, hmm but I get the same error. 2021-10-10 14:41:29 i think 2021-10-10 14:41:34 or you are missing the XDG directory 2021-10-10 14:41:47 sway/wayfire have VERY poor reporting as to why it fails 2021-10-10 14:41:50 it would shout about XDG 2021-10-10 14:42:01 psykose: I have that. 2021-10-10 14:42:07 $ XDG_RUNTIME_DIR=/tmp sway 2021-10-10 14:42:20 try that, humour me... I had it not reporting XDG issues 2021-10-10 14:42:31 and it was reporting some raster error instead 2021-10-10 14:42:48 mesa-dri-gallium ? 2021-10-10 14:43:37 psykose: I'm downloading that, good sign. 2021-10-10 14:43:58 It worked!! 2021-10-10 14:44:13 :) 2021-10-10 14:44:20 Actually my first time running wayland 2021-10-10 14:47:12 should just work assuming no weird driver issues 2021-10-10 14:54:22 How can I exit the session? How can I run programs? 2021-10-10 14:56:00 "Actually my first time running..." <- Or, no I used fedora before. 2021-10-10 14:56:38 Lookup a Sway tutorial 2021-10-10 14:57:09 On Xorg I use https://github.com/bakkeby/dwm-flexipatch 2021-10-10 15:01:47 super+shift+e is exit by default 2021-10-10 15:02:05 to run things you want a program runner like bemenu or whatnot bound to something 2021-10-10 15:04:06 i use `j4-dmenu-desktop --dmenu=bemenu` myself 2021-10-10 15:05:08 i don't remember if it copies over the config by default, you want a .config/sway/config copied from /etc/sway/config and then you can modify it with whatever 2021-10-10 15:05:19 the default one has comments and whatnot in it that should show you how some things work 2021-10-10 15:28:58 Good, I made my way into a terminal 2021-10-10 15:29:04 On Sway 2021-10-10 15:29:39 I pressed every button I deemed to be useful for that purpose. 2021-10-10 15:29:53 super-enter for terminal 2021-10-10 15:30:39 Yes, thanks 2021-10-10 15:54:47 Ok so, I know how to start waydroid. I only need binder_linux now and I should then be fine. PureTryOut (matrix.org) is Waydroid guaranteed to work on ARM64? 2021-10-10 15:55:07 Yes I've used it on my PinePhone 2021-10-10 15:55:08 * on ARM64 hardware? 2021-10-10 15:55:18 Good news 2021-10-10 15:55:19 Why are you so afraid that it won't support arm64 lol 2021-10-10 15:55:35 i don't think someone would have added it to packages if it didn't 2021-10-10 15:55:41 Arm64 on the desktop is rare stuff. 2021-10-10 15:55:56 maybe for games, everything else works fine 2021-10-10 15:56:08 everyone uses x86_64. 2021-10-10 15:57:37 50% of my desktops are arm64 ;) 2021-10-10 15:57:49 1/2? :p 2021-10-10 15:57:56 yes :D 2021-10-10 15:58:07 when counting desktops, if you measure time the split is quite different 2021-10-10 15:58:44 i wish there was more arm hardware 2021-10-10 15:59:20 everything is either some shitty 1gb soc, or the solidrun thing, or $10k for an altra system i guess 2021-10-10 16:00:05 yep 2021-10-10 16:00:10 my second desktop is a pi4 2021-10-10 16:00:15 2gb ram, but it's troublesome 2021-10-10 16:00:21 i guess you can run a server on the m1 mac already, but the desktop has some way to go 2021-10-10 16:01:35 Habbie: You connected your phones to your monitors? With arm64 you must mean raspberry pi's. 2021-10-10 16:01:46 errweas[m], yes, pi4, see a few lines up :) 2021-10-10 16:02:57 Sry, I was reading a message, replying and so on. 2021-10-10 16:04:00 ack :) 2021-10-10 16:04:23 * rare stuff you know. 2021-10-10 16:06:00 fyi matrix edits don't work super well over irc, as they just make a tiny diff with * in a regular message 2021-10-10 16:10:24 errweas: desktops sure but Waydroid is mainly focused on phones, which are a lot arm64 2021-10-10 16:10:26 thats because matrix is irc in easy mode. in irc-mode you gotta get it right first time 2021-10-10 16:11:41 it's mostly just funny when someone doesn't know how it sends then edits their message like 14 times 2021-10-10 16:12:15 bonus points if it's multiline so you get 14 long urls posted for each edit 2021-10-10 16:16:17 irc has a command to help with that... /kick 2021-10-10 16:18:10 hardly friendly 2021-10-10 16:18:53 debatable, one could say such spam isn't friendly and also goes against netiquette 2021-10-10 16:19:27 psykose, I know because I'm on matrix gitter rooms, each edit sends me a notification and people edit 10+ times 2021-10-10 16:19:36 :) 2021-10-10 16:19:39 :) 2021-10-10 16:20:01 Naib: people on bridges don't usually know how the features interact 2021-10-10 16:20:02 I love cross-platform messaging 2021-10-10 16:20:29 you just tell them once and they stop doing it 2021-10-10 16:21:17 thats a give, netiquette would equally expect an explanation/warning before corrective action is applied :) 2021-10-10 16:21:46 I always blame people responsible for said bridges 2021-10-10 16:21:59 if they bridge different platforms, it's on them 2021-10-10 16:22:45 anyway.. back to extracting seaborn html color codes to update waybar :) 2021-10-10 16:23:10 "somebody re-write the linux kernel in rust plz" rust people at it again, dammit 2021-10-10 16:23:38 where? I need some :popcorn: 2021-10-10 16:23:55 #rust:matrix.org 2021-10-10 16:36:21 id love to see linux rewritten in rust 2021-10-10 16:36:27 purely to piss off unix weenies 2021-10-10 16:36:46 1:1 rewrite or just rust unix? 2021-10-10 16:36:52 the fallout from that occuring would be hilarious 2021-10-10 16:36:56 because there is RedoxOS 2021-10-10 16:36:57 PJ[m]: 1:1 2021-10-10 16:37:32 well, 1:1 requires from Rust to mature on other architectures/platforms 2021-10-10 16:37:33 any significant portion of linux being rewritten in rust is enough to piss people off 2021-10-10 16:39:12 mushrooms are the fruiting body of large underground mycelial networks 2021-10-10 16:39:13 rewrite RedoxOS in C 2021-10-10 16:39:26 Rust is the fruiting body of a large ideology that exists below the surface 2021-10-10 16:39:35 i love mushrooms 2021-10-10 16:39:39 i don't 2021-10-10 16:39:43 theyre so good 2021-10-10 16:41:02 what ideology, valerius 2021-10-10 16:42:10 the Rust ideology, which is to put Rust components into anything and everything for the sole purpose of making the language popular, not for any real technical reasons 2021-10-10 16:43:32 Rust just happens to be the lucky language in this case, but the ideology is evangelism 2021-10-10 16:43:58 well, just to piss all unix and rust people, I will rewrite Linux in Go 2021-10-10 16:44:01 evangelism isn't exactly an ideology 2021-10-10 16:44:09 "well, 1:1 requires from Rust..." <- Btw, I compiled Veloren ( a voxel based game, written in rust) on ARM64. 2021-10-10 16:44:54 arm64 linux is pretty much first tier support so I wouldn't call that an achievement 2021-10-10 16:45:10 For me it is. 2021-10-10 16:46:10 I love Veloren veloren.net 2021-10-10 16:47:57 Hmm, I tried x86_64 emulation on Gentoo: segfaults and Ubuntu: no segfaults but every program has different issues. 2021-10-10 16:50:07 * different issues and fails because of them. 2021-10-10 16:54:33 PureTryOut (matrix.org) Do I have to put modules via akms into the Alpine Kernel or can I put in needed modules on other Distros and then transfer that Kernel over to Alpine? 2021-10-10 16:55:59 Uh what? For Waydroid? You don't need extra modules, just a cmd option for the kernel command line 2021-10-10 16:58:58 Just add `psi=1` to your kernel commandline, reboot, and it'll work 2021-10-10 17:01:46 Now that's awesome! 2021-10-10 18:01:31 Hmm, I enabled the `psi=1` kernel command in /etc/default/grub, updated grub and rebooted, same errors in waydroid log. 2021-10-10 18:02:14 Ahat does psi=1 have to do with waydroid anyway? psi is pressure stall information. 2021-10-10 18:02:32 s/Ahat/What/ 2021-10-10 18:02:56 s/Hmm, I enabled the `psi=1` kernel command in /etc/default/grub, updated grub and rebooted, same errors in waydroid log./PureTryOut (matrix.org) I enabled the `psi=1` kernel command in /etc/default/grub, updated grub and rebooted, same errors in waydroid log./ 2021-10-10 18:25:15 Waydroid tries to restart itself load ashmem, remount stuff and fails and tries again with a countdown from 10 to 1. 2021-10-10 18:25:51 s/Hmm, I enabled the `psi=1` kernel command in /etc/default/grub, updated grub and rebooted, same errors in waydroid log./PureTryOut (matrix.org) I enabled the `psi=1` kernel command in /etc/default/grub, updated grub and rebooted, same errors in waydroid log./ 2021-10-10 18:27:30 Is anyone using thttpd with alpine now? I am interested if there are problems with throttle files. I discovered what I think is a bug, fixed it, emailed Jef Poskanzer (thttpd author), no answer yet. 2021-10-10 19:03:48 "Just add `psi=1` to your..." <- PureTryOut (matrix.org) I enabled the `psi=1` kernel command in /etc/default/grub, updated grub and rebooted, same errors in waydroid log. 2021-10-10 19:04:06 What does psi=1 have to do with waydroid anyway? psi is pressure stall information. 2021-10-10 19:05:29 Yes, Android needs it for a userspace oom killer 2021-10-10 19:05:42 Oh I see. 2021-10-10 19:06:10 I can the waydroid log here if you want to 2021-10-10 19:06:14 For actual Waydroid support please ask in #waydroid:connolly.tech 2021-10-10 19:06:24 Ok 2021-10-10 20:55:01 I think there isn't a wiki article on installing and running virtual machines with alpine as the host. Are the best practices covered on the KVM article? Is KVM the way I should go? 2021-10-10 20:58:24 qemu + kvm is quite common 2021-10-10 20:58:45 maybe with libvirt if you want some higher level control 2021-10-10 21:05:50 yeah, some form of handholding is gonna have to be necessary, i think 2021-10-10 21:06:17 is it as "simple" as how vbox just lets me specify a bunch of parameters and get windows booted? was just going to fuck around with that and some other operating systems today 2021-10-10 21:07:39 ime with qemu you just run 2 commands and windows worked 2021-10-10 21:07:42 but it's been a while 2021-10-10 21:07:51 that's fuckin neat 2021-10-10 21:07:51 man 2021-10-10 21:08:04 so much work already done for me 2021-10-10 22:20:03 Can someone explain the correlation between nftables and iptables? What's the NEW linux firewall stuff? Does it all use the same shit on the backend now? 2021-10-10 22:23:50 And WTF is netfilter? 2021-10-10 22:27:14 well, their homepage explains it pretty well imho 2021-10-10 22:27:41 https://www.netfilter.org/ 2021-10-10 22:28:26 But does it relate itself to everything else? 2021-10-10 22:28:35 ah there it is 2021-10-10 22:29:00 does nftables have an nf6tables thing 2021-10-10 22:29:17 nope, i was wrong, lmao; that's on the page, too 2021-10-10 22:29:19 thanks, PJ 2021-10-10 22:29:29 👍 2021-10-10 22:29:36 really rare that i see the homepage of a project give the information i really care about, except for on linux stuff, it seems 2021-10-10 22:29:46 it's almost like all the people making stuff for linux USE LINUX 2021-10-10 22:30:04 whereas half the github project pages i read are just mac fuckboys trying to make it to the hn frontpage 2021-10-10 22:30:08 I feel you, I get angry when I can't find link to git repo 2021-10-10 22:30:20 since that is mostly what I care about 2021-10-10 22:30:48 I know few of those that are not 'mac fuckboys' 2021-10-10 22:30:57 i do think that's an interesting thing i currently lack; just being able to look at how something works by reading the code 2021-10-10 22:31:29 since i'm not familiar enough with linux dev to know when certain libs are used for certain purposes 2021-10-10 22:32:17 oh well, I don't know much of linux either, but I know and have skimmed through quite few programming languages to get general idea what it does 2021-10-10 22:32:25 works good enough for me 2021-10-10 22:32:38 that's emboldening; i'll keep it in mind 2021-10-10 22:33:06 except for C, god forbids me from understanding that elvish language 2021-10-10 22:33:12 ikke: are you on matrix, too? at what point is keeping this tied to an irc even worth it? kinda funny how i oppose anything regarding discord or other chatsites, but matrix works just fine in its stead, in my head, at least 2021-10-10 22:34:17 matrix still lacks good GUI client with proper UX, Element is something but whenever I use it it, I feel meh 2021-10-10 22:34:35 and lacks proper administration 2021-10-10 22:34:39 fluffychat.. 2021-10-10 22:34:40 * proper administration via client 2021-10-10 22:34:46 fluffychat is terrible 2021-10-10 22:34:51 but that doesnt work on musl 2021-10-10 22:34:55 ive never had issues 2021-10-10 22:35:01 It's even worse than Element imfo 2021-10-10 22:35:12 granted, i use it primarily on mobile 2021-10-10 22:35:17 It's like someone developed a very old Android app and then ported it to desktop 2021-10-10 22:35:29 it is an android app first and foremost 2021-10-10 22:35:33 thats what flutter is 2021-10-10 22:35:36 it's also desktop 2021-10-10 22:35:45 yea 2021-10-10 22:36:03 still, cant downplay the value of having a very good mobile app 2021-10-10 22:36:16 i switched to cinny recently, PJ 2021-10-10 22:36:17 Element works fine for me on mobile, I actually don't have anything bad to say about it except for strict IRC theme like on desktop and all features that Element Desktop lacks 2021-10-10 22:36:33 but would prefer if the lazy fucks behind weechat-matrix-rs would actually finish it 2021-10-10 22:36:34 I have Chinny running in Docker, it's still basically feature-less 2021-10-10 22:36:42 yeah, lol 2021-10-10 22:36:47 why are matrix projects so fucking slow 2021-10-10 22:36:49 fluffychat has almost full feature parity w element 2021-10-10 22:36:54 complex protocol 2021-10-10 22:36:56 BUT, it is 100000% looking better than Element 2021-10-10 22:36:58 unneccessarily so 2021-10-10 22:36:59 it seems like all of them are linked together in being real slow 2021-10-10 22:37:01 i imagine 2021-10-10 22:37:07 but thats how it is 2021-10-10 22:37:31 Whenever I see Chinny I instantly feel better and don't want to puke, like the client just looks good and approachable 2021-10-10 22:37:52 yeah, that's the nice feeling i get, too, as i'm not raging over the slowness of element 2021-10-10 22:38:02 element is chug city 2021-10-10 22:38:03 I keep saying 'Chinny', pj, stop... 2021-10-10 22:38:08 lmao 2021-10-10 22:38:11 got any cinny screenshots? 2021-10-10 22:38:16 never even heard of it 2021-10-10 22:38:29 i just use element in a browser and its fast enough for me 2021-10-10 22:38:47 cinny looks very similar in the browser for me, as well, but a bit more typographically sane 2021-10-10 22:38:53 my frame of reference is discord however 2021-10-10 22:38:59 ACTION uploaded an image: (618KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/uSWUWoTKMfpymNcPIddTFUAl/image.png > 2021-10-10 22:39:03 i fucking despise discord 2021-10-10 22:39:09 yeah that's what mine looks like right now as well 2021-10-10 22:39:11 cf pj 2021-10-10 22:39:23 element was lightning fast on my celeron compared to fucking discord 2021-10-10 22:39:35 ding ding ding, Discord is electron 2021-10-10 22:39:41 100% true; the new gen of users is really fucked over in terms of UI 2021-10-10 22:39:42 no fucking wonder 2021-10-10 22:39:43 i was using it in the browser 2021-10-10 22:39:49 that's still Electron 2021-10-10 22:39:54 hm? 2021-10-10 22:40:04 like the main engine behind it is still heavy 2021-10-10 22:40:07 ye 2021-10-10 22:40:08 yea 2021-10-10 22:40:14 irs a quality of webapps 2021-10-10 22:40:18 no matter if it's desktop app or browser 2021-10-10 22:40:31 miracle how fucked up discord is though 2021-10-10 22:40:52 Revolt is great Discord alternative 2021-10-10 22:40:55 like electron apps can be slightly less bad with a bit of effort 2021-10-10 22:41:09 discord takes this effort 2021-10-10 22:41:12 magnifies it 2021-10-10 22:41:19 and makes it worse 2021-10-10 22:41:30 I keep saying that and will probably be always, VSCode from Microsoft is the only Electron app that does not suck and isn't slow 2021-10-10 22:41:52 vsc is the exception that proves the rule 2021-10-10 22:41:57 although they keep using old Nodejs/Electron 2021-10-10 22:42:25 and nodejs is a bitch to get working on musl 2021-10-10 22:42:39 when i've used vs code, it wasn't as snappy as emacs or vim \:^) 2021-10-10 22:42:40 * and electron/nodejs, * nodejs/v8/chromium is 2021-10-10 22:43:22 listen, I can use magnetic needle and 0 RAM/CPU to code shit 2021-10-10 22:43:30 ;) 2021-10-10 22:44:33 there are languages out there designed to be capable of being programmed outside of a computer 2021-10-10 22:44:38 theyre cool 2021-10-10 22:45:04 I've written C++ on a piece of paper year ago, isn't hard 2021-10-10 22:45:26 makes you really self-aware to not typo 2021-10-10 22:45:38 its considerably more annoying to write c++ on paper than it is to write apl :p 2021-10-10 22:46:38 languages where you write less are better suited for environments where writing is more annoying 2021-10-10 22:48:04 granted, theyre a bit inscrutable 2021-10-10 22:48:31 i write psuedocode on paper and traditional languages nake it annoying, haha 2021-10-10 22:48:38 i cant imagine java on paper 2021-10-10 22:48:39 god 2021-10-10 22:49:00 rust would be fun on paper 2021-10-10 22:49:56 burroe checker!! 2021-10-10 22:50:50 the day when python dies is when I'll achieve eternal peace 2021-10-10 22:52:45 the fact that somewhat verbose problems are oneliners is a nice advantage for writing on paper 2021-10-10 22:53:17 like, fib is `{1_x(|+\)/!2}` 2021-10-10 22:53:18 writing problems on paper is just a different kind of rubber ducky 2021-10-10 22:54:12 all non unique characters of a string is `&1<#'=:` 2021-10-10 22:54:37 complete noise to anyone who doesnt know the language 2021-10-10 22:54:42 but easily transferrable 2021-10-10 22:55:51 abolish strings, write everything using ints 2021-10-10 22:56:01 ah 2021-10-10 22:56:04 but you see 2021-10-10 22:56:10 strings are just ints 2021-10-10 22:56:14 in a row 2021-10-10 22:56:30 in a slice 2021-10-10 22:56:47 whats a slice 2021-10-10 22:57:01 fancy name for an array 2021-10-10 22:57:09 oh yea 2021-10-10 22:57:14 except slice doesn't have defined size 2021-10-10 22:57:18 it varies based on languages 2021-10-10 22:57:31 funnily enough, array languages dont have slices 2021-10-10 22:58:13 no references 2021-10-10 22:58:33 I mean, I could be using C# types but no one would probably get them 2021-10-10 22:58:43 go on 2021-10-10 22:58:49 ArrayList 2021-10-10 22:59:01 lol 2021-10-10 22:59:15 terrible 2021-10-10 22:59:30 this is why i hate types 2021-10-10 22:59:54 I mean, C# does have Array 2021-10-10 23:00:08 but normally you don't use those because they are weak 2021-10-10 23:00:49 C# is a fun language, truly 2021-10-10 23:02:13 PJ[m]: puny array vs stronh arraylist? 2021-10-10 23:02:17 tf is a weak type 2021-10-10 23:03:41 I mean, ArrayList is not recommended either now 2021-10-10 23:03:51 we are supposed to use `List` 2021-10-10 23:04:24 it all comes down from what underlying types are used 2021-10-10 23:05:27 depends on those types, different functions are provided 2021-10-10 23:09:50 wow, WOW, really Element, now you think my password is valid? I've been trying for few days and it's correct today 2021-10-10 23:11:22 "tf is a weak type" <- it just sucks because it's slow and old 2021-10-10 23:42:06 can i do shit like "hot-plug all of the devices in the front usb panel to the guest" with qemu? 2021-10-11 06:41:25 Hi. I would like to disable the switch to framebuffer mode happening as part of the boot sequence (this is to control the resolution chosen during installation), but on my devices the switch cleans the monitor. Could you help me towards the component doing this? 2021-10-11 08:02:21 What do y'all use for downloading and building custom kernels? Just git or is there some convenient wrapper? 2021-10-11 08:39:01 Do I see it correctly, that static packages (e.g. libzmq-static) are statically linked to libstdc++ libgcc etc.? So shouldn't the resulting license then be GPL and not LGPL with link exception? 2021-10-11 10:21:44 Is there a way to customize how apk installs files in /etc and similar directories? 2021-10-11 10:21:54 ccx^xmpp: no 2021-10-11 10:22:13 What do you want to customize? 2021-10-11 10:25:27 Put in custom versioning and/or configuration management integration. 2021-10-11 10:25:58 apk will not overwrite files in /etc if they are modified 2021-10-11 10:26:26 at most it will create .apk-new files 2021-10-11 10:29:14 Does moving the file and symlinking count as modification? 2021-10-11 10:29:47 I don't expect it to overwrite symlinks 2021-10-11 10:37:52 Is there a way to list which packages have post-install/update hooks? 2021-10-11 10:40:23 I don't think so at the moment' 2021-10-11 10:56:46 would it be a good idea for ISPs to have/promote XMPP ? 2021-10-11 10:57:14 or i mean webhotels... like they do with mail 2021-10-11 11:06:12 no? 2021-10-11 11:12:17 why? 2021-10-11 11:12:54 because it's not secure by default 2021-10-11 11:13:36 also it's quite dead 2021-10-11 11:14:22 maybe an alternative to it then ... for calls/messaging, independent of centralized big tech 2021-10-11 11:15:12 jitsi/matrix/rocketchat to name a few are quite popular better alternatives 2021-10-11 11:15:44 Revolt if you want Discord UX 2021-10-11 11:16:44 XMPP would only make sense with OMEMO but clients for it after all this years are still shite 2021-10-11 11:21:36 ah thought xmpp was more secure when set up correctly, and more lightweight/flexible, dont really know tho 2021-10-11 11:21:48 is jitsi/rocketchat protocoals? 2021-10-11 11:24:15 every single client i've touched supports omemo fine 2021-10-11 11:24:26 jitsi uses webrtc, don't know what rocketchat uses since I haven't used it myself (yet) 2021-10-11 11:25:19 probably something built on top of HTTPS/WebSockets 2021-10-11 11:25:28 https://github.com/RocketChat 2021-10-11 12:10:41 CcxWrk: what do you mean 2021-10-11 12:11:37 darkdragon: on unix-like systems, static libraries do not contain their dependencies. somewhat unfortunately, static libraries also do not contain a list of their dependencies 2021-10-11 14:05:47 Hello71: Thanks for your insights. When build an executable with such a library, I need the original dependencies as well or can I link to other dependencies also? I mean linking to libc++ instead of libstdc++ doesn't work, right? 2021-10-11 14:09:32 hey there 2021-10-11 14:09:41 has anyone got https://github.com/luvit/luv working on alpine? 2021-10-11 14:09:51 I'm trying to get it working in a docker container 2021-10-11 14:10:05 but I'm getting 2021-10-11 14:10:09 "error relocating /usr/lib/libluv" 2021-10-11 14:15:18 i just installed it with luarocks and it went fine 2021-10-11 14:15:25 what steps are you taking 2021-10-11 14:34:58 XMPP has good secure clients everywhere but iOS (because GPL incompatibility of the App Store) 2021-10-11 14:36:14 Definitely not dead. It's hard to measure private messengers though, as conversations tend to be, well, private. 2021-10-11 14:37:15 it's actually very easy, you can measure it by seeing who uses it and how it develops over time 2021-10-11 14:39:07 I can't tell anything about iOS/macOS since I don't use them but both Linux/Windows clients have not progress much in past 10+ years 2021-10-11 14:41:35 Dino had 0.2 release Nov 2020 2021-10-11 14:43:02 not sure what there is to progress 2021-10-11 14:43:20 i have not noticed any improvement in IM usability since the msn days 2021-10-11 14:43:27 everything is the same shit reskinned with bugs shuffled around 2021-10-11 14:43:35 https://github.com/dino/dino/pulse/monthly 2021-10-11 14:43:37 it's pretty active 2021-10-11 14:44:12 and every once in a while, someone starts something new, and everyone thinks it's some new great thing but it just follows the same process of "starts minimal" -> "too many features, sucks" 2021-10-11 14:44:15 rinse repeat 2021-10-11 14:44:36 commits are not something you should count if the project is alive 2021-10-11 14:44:43 psykose: ok I'm started, now what? :-) 2021-10-11 14:44:56 venture forth :) 2021-10-11 14:45:15 Forth? I stopped using that in the 1980s lol 2021-10-11 14:45:22 :p 2021-10-11 14:46:04 not sure what are you on about, jabber lacked security, many companies built on top of jabber most notably Facebook Messenger was XMPP 2021-10-11 14:46:30 They had good UX since it's a company with big money 2021-10-11 14:46:42 what messenger around the time xmpp started had good security 2021-10-11 14:46:56 gpg doesn't count 2021-10-11 14:47:13 didn't knew gpg is a messenger :) 2021-10-11 14:49:00 that's obviously not what i mean 2021-10-11 14:49:02 so far Discord is pretty much the best UI/UX wise 2021-10-11 16:01:47 Using 3.14, how do I get the udhcpc6 to work in getting dhcpv6 functioning? 2021-10-11 16:08:55 should be like "iface eth0 inet6 dhcp", no? 2021-10-11 16:10:24 I thought so too, but apparently that doesn't kick off the executable in the same way that "iface eth0 inet dhcp" does. 2021-10-11 16:24:51 seems like ifupdown-ng doesn't support udhcpc6, not sure why 2021-10-11 16:25:00 you can install dhcpcd or dhclient 2021-10-11 16:25:13 actually i'm not sure dhclient support dhcpv6, but i know dhcpcd does 2021-10-11 16:25:19 https://github.com/ifupdown-ng/ifupdown-ng/blob/master/executor-scripts/linux/dhcp 2021-10-11 16:36:39 `iface eth1 inet6 auto` 2021-10-11 16:38:03 it's mentioned on alpine wiki 2021-10-11 16:38:49 https://wiki.alpinelinux.org/wiki/Configure_Networking#IPv6_Stateless_Autoconfiguration 2021-10-11 16:38:54 i don't think that actually does anything 2021-10-11 16:39:11 as soon as you bring up the interface, kernel will do ra 2021-10-11 16:39:22 and neither busybox ifupdown nor ifupdown-ng invoke udhcpc6 2021-10-11 16:40:10 even the article says it's for slaac, which is half-true 2021-10-11 16:40:32 works for me though 2021-10-11 16:41:38 ... 2021-10-11 16:41:58 it does dhcpv6? or you just assume it does because you don't know what dhcpv6 means 2021-10-11 16:42:56 ... 2021-10-11 17:03:41 Is there a simple cookbook on how to compile a package for alpine on another system? Debian/devuan? 2021-10-11 17:04:32 user____: https://gitlab.alpinelinux.org/alpine/docker-abuild for example 2021-10-11 17:06:42 thanks 2021-10-11 17:07:20 Oh, that runs the build in an alpine container ikke 2021-10-11 17:07:35 I need something more bare metal. Since I am not set up for docker containers 2021-10-11 17:07:49 You can setup alpine in chroot 2021-10-11 17:08:01 right 2021-10-11 17:08:03 that's the other option 2021-10-11 17:08:07 https://wiki.alpinelinux.org/wiki/Alpine_Linux_in_a_chroot 2021-10-11 17:09:31 bookmarked both, thanks 2021-10-11 17:10:26 When a standard alpine package is rebuilt from source, are there any special steps to take? Specifically, thttpd-2.29 which I patched as I need. 2021-10-11 17:11:09 user____: You can add /home//packages/ to /etc/apk/repositories 2021-10-11 17:11:22 make sure you add your public key to /etc/apk/keys 2021-10-11 17:11:26 and then you can install it 2021-10-11 17:11:30 also make sure you bump pkgrel 2021-10-11 17:12:17 Okay, another set of things to write up. 2021-10-11 17:13:25 Wait, this is for distributing a rebuilt package, right? Not for installing it on a single alpine host? 2021-10-11 17:13:56 maybe it would be better if you describe what exactly you are trying to achieve? 2021-10-11 17:14:29 you want to rebuild a thttpd on a separate host for an alpine host? 2021-10-11 17:15:01 but not necessarily update the package in alpine repos? 2021-10-11 17:15:18 yes. I patched and tested the patched thttpd on a Devuan host and now I would like to put it in the vm I have here which is an alpine vm 2021-10-11 17:15:22 PJ[m]: exactly. 2021-10-11 17:16:33 What compiler is used to build userland alpine? gcc or... ? 2021-10-11 17:16:47 gcc 2021-10-11 17:17:29 You can somewhat follow https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package and when you build it you should be able to copy the `~/packages` into that host, then follow the step `sudo apk add --repository /home/USER/packages/testing $pkgname` 2021-10-11 17:17:34 Ok. Does it matter a lot if I'm not using the latest version for testing? My local vm is based on 'alpine-extended-3.14.1-x86_64' 2021-10-11 17:18:00 PJ[m]: great, I think that info was exactly what I needed. 2021-10-11 17:18:16 user____: then you need to make sure the chroot you build the package with is also using alpine 3.14 2021-10-11 17:18:20 not edge 2021-10-11 17:18:25 sure 2021-10-11 17:18:47 Are there large jumps in *libc etc versions between 3.14 and edge etc?? 2021-10-11 17:18:49 s|/home/USER/packages/testing|/home/USER/packages/branch| 2021-10-11 17:19:41 ok, I'll be back later, thanks for now 2021-10-11 17:19:50 user____: atm only patches 2021-10-11 17:33:06 Hello71: Sorry, missed most of this dialog when I went to lunch. And yes, stateless autoconfig works just fine as is. It's proper stateful DHCPv6 that doesn't seem to work. It's clear that the busybox binary is there, but the scripts to support it don't exist yet. 2021-10-11 17:33:24 dhcpv6 isn't very popular 2021-10-11 17:34:41 Except in enterprise environments 2021-10-11 17:42:20 yes, also useless proxies that require HTTP CONNECT for HTTPS 2021-10-11 17:42:25 defeating the whole purpose of a proxy 2021-10-11 17:43:31 Well, also in environments with DDNS, and where only the DHCP server is authorized to send the DDNS updates. 2021-10-11 17:43:49 Without DHCPv6, those updates aren't possible. 2021-10-11 17:48:16 so... instead of trusting the clients to send their hostname, you instead trust the dhcp server to send the hostname... that it got from the client? 2021-10-11 17:48:23 Correct. 2021-10-11 17:49:09 The question of trust is more than just what hostname it has. It's also the proper aging out of withdrawn leases, which should not be the responsiblity of the client. 2021-10-11 17:50:51 why can't the dns server handle this itself 2021-10-11 17:51:16 What do you mean? How would a DNS server be prompted to update itself? 2021-10-11 17:51:33 so that the dhcp server doesn't need to keep a persistent list of leases, and if it ever loses track of a single one, the dns server will be clogged forever 2021-10-11 17:52:31 It's not the role of DNS to keep track of address assingments. It's for DNS to keep track of name to address records, which may or may not be attached to a lease. That's what DHCP is for. 2021-10-12 02:18:15 Hello! I've been using Alpine Linux as my desktop OS on my laptop and the battery died while updating. Right now it boots up the partition as readonly, hostname looks like it is gone, and on boot it says it is unable to create /run/openrc. Is there any way I can recover this? I have booted up the installation media now, and as simple as Alpine is, it seems like I could replace some files from the 2021-10-12 02:18:35 installation media to get it to boot right but I am not quite sure since I haven't done it before. 2021-10-12 02:24:28 Back online, not sure what happened... 2021-10-12 02:33:27 Going to experiment and see what happens... 2021-10-12 04:00:08 I guess I'm not making very good progress... Seems like there should be an easy way to repair an installation 2021-10-12 04:52:51 No clue if this is bad advice; what if you boot live media and then pull stuff over for the chroot? 2021-10-12 04:53:03 Were you on stable or edg--he's gone. 2021-10-12 06:01:05 Hi, it seems that the Alpine virt kernel doesn't have sound support? 2021-10-12 07:07:21 that is correct 2021-10-12 07:12:23 PJ[m]: Why is that? 2021-10-12 07:13:36 It doesn't include drivers, as for reason, virt kernel is mostly used for cloud providers where there is no sound card 2021-10-12 07:15:25 PJ[m]: So if I want sound support on my VM I need to compile my own kernel? 2021-10-12 07:15:52 You can just use standard kernel 2021-10-12 07:16:41 Ok, thanks 2021-10-12 09:02:10 https://reclaimthenet.org/firefox-ads-address-bar/ 2021-10-12 09:06:06 frag: https://gitlab.alpinelinux.org/alpine/aports/-/issues/13068 2021-10-12 09:08:25 PJ[m]: very cool :) 2021-10-12 09:09:58 will there be an "alpine news magazine" :P 2021-10-12 09:10:57 isn't that just https://alpinelinux.org/ ;) 2021-10-12 09:12:27 hehe 2021-10-12 13:48:37 What is the rationale for installing some terminfo definitions in /etc/terminfo and some in /usr/share/terminfo? Wasn't /etc/terminfo supposed to be site-local overrides? 2021-10-12 15:16:08 Can anyone point me to an aport that is apkbuild-pypi compatible so I can try and figure out how to make a new aport for this type of project? 2021-10-12 15:22:45 as far as i understand reading it, it's mean to make an aport in the first place, so it's not reverse dependent on some aport already. you use it once 2021-10-12 15:23:41 Oh, okay. So it should make a suitable template for me to edit 2021-10-12 15:23:44 you pass it some pypi package and it fills out the default fields on getting stuff from there 2021-10-12 15:23:46 yea more or less 2021-10-12 15:23:49 Alright, I guess I can start there 2021-10-12 15:23:53 Thanks, again 2021-10-12 15:57:10 helloo, i have sound working with alsa (tested with 'speaker-test'), but audio in firefox is missing. i guess pulseaudio is needed, but 'rc-service pulseaudio start' tells me it should not be syarted systemwide. 2021-10-12 15:57:27 i wonder how i should fix that? 2021-10-12 15:57:32 it shouldn't, it gets launched on-demand 2021-10-12 15:58:11 you can add something like: start-pulseaudio-x11 to some .xinitrd file 2021-10-12 15:58:23 or go straight for pipewire 2021-10-12 15:59:46 ah ok 2021-10-12 16:00:47 i am using sway currently. so would i still use start-pulseaudio-x11 or something? 2021-10-12 16:00:59 ha ... good question :) 2021-10-12 16:01:03 just running `pulseaudio` should work 2021-10-12 16:01:05 or whatever it was called 2021-10-12 16:01:07 im using sway, but i am also using pipewire 2021-10-12 16:01:19 ah 2021-10-12 16:01:21 it will abort if it's already started afaik 2021-10-12 16:01:25 just run em and see 2021-10-12 16:01:41 pulseaudio -D makes it bg itself if my memory serves 2021-10-12 16:01:54 abort in a ZOMG I can't lock socket as oppose to "I am written as a singleton and there is already a server running" 2021-10-12 16:02:16 that's basically the same thing 2021-10-12 16:03:12 true, except the error messages are really not clear to indicate its because something else was forfilling that need 2021-10-12 16:03:47 if i remember correctly the error does say something like 'is another server running?' 2021-10-12 16:04:29 maybe 2021-10-12 16:15:46 Naib: why did you decide to use pipewire instead of for instance pulseaudio? 2021-10-12 16:20:40 abuild is always looking one level higher for the build files than they get extracted to. What am I configuring wrong from the template? 2021-10-12 16:20:47 source="https://gitlab.xfce.org/apps/$pkgname/-/archive/$pkgname-$pkgver/$pkgname-$pkgname-$pkgver.tar.gz" 2021-10-12 16:20:47 builddir="$srcdir/" 2021-10-12 16:21:16 [~/Sources/aports/testing/catfish]$ ls src/ 2021-10-12 16:21:16 catfish-catfish-4.16.3 catfish-catfish-4.16.3.tar.gz 2021-10-12 16:23:36 don't you want $srcdir/$pkgname-$pkgname-$pkgver 2021-10-12 16:23:57 you should probably rewrite file name 2021-10-12 16:24:48 `source="filename.tar.gz::link"` 2021-10-12 16:30:56 psykose: So I'm meant to modify the srcdir line to reflect how the sources get downloaded/named? 2021-10-12 16:31:10 builddir 2021-10-12 16:31:33 PJ: So that first part before the double colon defines the final name of what gets fetched? 2021-10-12 16:31:39 Yes, builddir, sorry 2021-10-12 16:31:55 yes, it renames the fetched file from link into the first part 2021-10-12 16:32:08 Ohhh, okay. Cool 2021-10-12 16:32:42 lemontree:yes 2021-10-12 16:33:13 It only affects the archive name, not the dirname of extracted files 2021-10-12 16:34:56 ikke: Yep, I'm just noticing that now. Extrcated folder is still catfish-catfish-4.16.3 and abuild is failing to find anything still... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/wXpIFxCnwMItEtPOjZOtyFhg) 2021-10-12 16:35:10 Saijin_Naib[m]: that's where you set builddir for 2021-10-12 16:35:45 should have worded that differently, but my suggestion is not linked to your problem :) 2021-10-12 16:35:51 So this is wrong?... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/kYthMTiConSmHrGVvVkBUdRz) 2021-10-12 16:36:02 most likely yes 2021-10-12 16:36:04 yea 2021-10-12 16:36:07 do the latter 2021-10-12 16:36:16 And this is normal? one normally works backwards like this to find the right builddir? 2021-10-12 16:36:25 it's just the name of the dir in src 2021-10-12 16:36:30 Or am I doing something wrong with the setup? 2021-10-12 16:36:36 no, you are correct 2021-10-12 16:36:45 but it's just the name of the dir when you untar the archive 2021-10-12 16:36:53 you should just `abuild checksum`, `abuild unpack` and then you can just `ls src/` to see what is unpacked 2021-10-12 16:37:06 Saijin_Naib[m]: What url do you use for source? 2021-10-12 16:37:47 https://gitlab.xfce.org/apps/catfish/-/archive/catfish-4.16.3/catfish-catfish-4.16.3.tar.gz 2021-10-12 16:38:04 I found it from their Repository->Tags page 2021-10-12 16:40:01 Right, then you need to do that indeed 2021-10-12 16:40:12 nothing wrong or unusual with it 2021-10-12 16:40:24 Oh, okay. Cool, thank you. 2021-10-12 16:40:24 they seem to have tarballs here as well: https://archive.xfce.org/src/apps/catfish/4.16/ 2021-10-12 16:40:42 Is it better to get it from their code hosting or from them? Same-ish thing? 2021-10-12 16:40:47 What's better practice? 2021-10-12 16:41:14 static archives are better, the automatically generated once can change (different sha512sum) 2021-10-12 16:41:24 ie, when the gitlab version gets updated 2021-10-12 16:42:48 Hmm, that makes sense 2021-10-12 16:45:49 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/akpFPoxapkjtyYjzfasoalHi 2021-10-12 16:46:02 Does this failure mean I need to add pytest and py and pluggy to my makedepends 2021-10-12 16:47:29 Sorry, checkdepends 2021-10-12 16:51:05 py3-pluggy py3-pytest py3-py yeah 2021-10-12 16:53:54 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/OSbpqissLtIsdAXjZFVDeFkd 2021-10-12 16:54:03 No change in behavior it seems 2021-10-12 16:55:26 ah no 2021-10-12 16:55:30 this package doesn't have any tests 2021-10-12 16:55:36 my mistake 2021-10-12 16:56:06 unless i cannot see them 2021-10-12 16:56:44 then it's just `options="!check" # No testsuite` 2021-10-12 16:58:28 Is that allowed? 2021-10-12 16:58:35 I guess obviously if you're telling me but haha 2021-10-12 16:59:00 not everything has tests 2021-10-12 16:59:22 Hmm... Should I comment/remove the checkdepends line too then? 2021-10-12 16:59:30 remove 2021-10-12 16:59:53 if it had tests, you could make a check() and call just pytest or whatever the correct form is for testing upstream 2021-10-12 17:00:13 Ahh, okay. So that whole check block can get dropped, too? 2021-10-12 17:00:22 yep, no tests 2021-10-12 17:00:30 if they ever add them you can throw them back in 2021-10-12 17:00:42 My install is empty set (install=''), should I drop that too? 2021-10-12 17:00:47 That sounds good 2021-10-12 17:02:15 install= is for install scripts 2021-10-12 17:02:20 https://github.com/alpinelinux/abuild/blob/master/APKBUILD.5.scd#install-scripts 2021-10-12 17:02:58 the things you add as pkg.post-install, etc, if you don't need to do some setup for the package you don't need any 2021-10-12 17:03:02 Ok. So like for everything else, if I don't need it, drop it instead of commenting out or leaving default 2021-10-12 17:03:17 yep 2021-10-12 17:03:40 Awesome 2021-10-12 17:03:58 Does this get an arch="all" or arch="noarch" 2021-10-12 17:04:03 Since it seems to be pure python 2021-10-12 17:04:12 Build log seems to suggest noarch 2021-10-12 17:04:43 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/exhWDcrMvIzJoCdqtKcSAQBe 2021-10-12 17:04:45 "Package architecture(s) to build for. Can be one or several seperated by whitespace of: x86, x86_64, armv7, armhf, aarch64, ppc64le, s390x, mips64, all, or noarch, where all means all architectures, and noarch means it's architecture-independent (e.g., a pure-python package)." 2021-10-12 17:05:05 https://wiki.alpinelinux.org/wiki/APKBUILD_Reference#arch 2021-10-12 17:05:30 Ah, right there in the green banner haha 2021-10-12 17:05:32 Thank you 2021-10-12 17:05:46 Now in terms of Alpine-ness, the $pkgname-lang thing... Don't do that, right? 2021-10-12 17:05:58 Build failed without intltool, so I had to add it to makedepends 2021-10-12 17:06:14 Do that = add to subpackages alongside $pkgname-doc 2021-10-12 17:06:40 -lang is for translation files 2021-10-12 17:06:51 Is that something generally that is wanted in Alpine? 2021-10-12 17:07:09 if it makes a /usr/share/locale/$LANG with various stuff in it, you want it yes 2021-10-12 17:10:47 Oh wow, it worked. Thanks so much! 2021-10-12 17:11:02 Now, so i don't give like, 30 people heart-attacks reading my commit message 2021-10-12 17:11:11 What would one say is a good commit message 2021-10-12 17:12:12 I see some new aports don't have a commit message, it seems 2021-10-12 17:12:15 0https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/COMMITSTYLE.md 2021-10-12 17:12:49 testing/catfish: new aport 2021-10-12 17:14:01 so from my terminal, that'd be a git add . && git commit -a && git push 2021-10-12 17:14:10 and the message in vi would be testing/catfish: new aport 2021-10-12 17:14:42 you can also do git add -A 2021-10-12 17:14:53 Is that better synatx? 2021-10-12 17:14:56 Okay 2021-10-12 17:14:58 I suggest git add -p (--patch) 2021-10-12 17:15:07 and verify what you stage / commit 2021-10-12 17:15:08 mmm 2021-10-12 17:15:17 Okay 2021-10-12 17:16:42 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/fxsngAAreRUVqKVPjPURYrmW 2021-10-12 17:17:10 Saijin_Naib[m]: yeah, it does not work for untracked files 2021-10-12 17:17:25 git status -uall 2021-10-12 17:17:43 APKBUILD 2021-10-12 17:18:16 git add APKBUILD 2021-10-12 17:18:18 so just git add that and then try again with ```git add -p && git commit -a && git push``` 2021-10-12 17:25:55 you can't use -p until the file is already committed 2021-10-12 17:26:02 just add commit push, doesn't matter here 2021-10-12 17:38:23 awesome 2021-10-12 17:38:35 In setting up my MR, I have my bad commits from lsd 2021-10-12 17:38:43 Can I hide those from the GitLab UI? 2021-10-12 17:40:11 if you mean they are part of your branch, just remove them 2021-10-12 17:40:41 https://gitlab.alpinelinux.org/Saijin-Naib/aports/-/tree/Saijin-Naib-catfish 2021-10-12 17:41:21 Revert? 2021-10-12 17:41:24 the diff between this and aports should be the single catfish commit 2021-10-12 17:41:32 It isn't ☹️ 2021-10-12 17:41:57 first undo the last commit 2021-10-12 17:42:07 git reset --soft HEAD~1 2021-10-12 17:42:26 then you have multiple ways of making those changes into a new branch 2021-10-12 17:42:34 done 2021-10-12 17:43:03 i would probably git stash, git checkout master, git pull (to update master), then make a new branch off master and git stash pop 2021-10-12 17:43:19 Alright, thank you 2021-10-12 17:43:54 I think maybe the problem is the remote is my fork not aports 2021-10-12 17:44:01 it should be your formk 2021-10-12 17:44:03 fork* 2021-10-12 17:44:13 you make the mr from your fork in the web interface afterwards 2021-10-12 17:44:19 and the mr is from a branch in your fork 2021-10-12 17:44:22 Yep, that's what I did last time 2021-10-12 17:44:25 indeed 2021-10-12 17:44:33 And I make a branch off master after my commit got pushed 2021-10-12 17:45:18 if this has all the previous commits that means you never made a new branch 2021-10-12 17:45:36 after something you make gets merged, just delete the branch you had that was part of the mr 2021-10-12 17:45:54 and anything you work on should be its own branch off master 2021-10-12 17:45:56 Oof... I think I commited to master and then branched from that in the UI to make the MR 2021-10-12 17:46:00 -_- 2021-10-12 17:46:03 I did this wrong from the outset 2021-10-12 17:46:21 that makes no difference, the branch still has the same commit in it 2021-10-12 17:46:48 though it makes it harder to change that commit 2021-10-12 17:47:15 so, usually wrong yes 2021-10-12 17:47:27 Alright 2021-10-12 17:47:31 I think it's nuke from orbit time 2021-10-12 17:47:40 Maybe it's best I do this from the web ui for now 🤣 2021-10-12 17:48:09 you can't do this from the web ui 2021-10-12 17:48:15 this is part of your local branch workflow 2021-10-12 17:48:21 before you do any pushing to anywhere 2021-10-12 17:48:34 Oh, okay. 2021-10-12 17:48:48 the issue was just in which branch you were working on locally, that's all 2021-10-12 17:49:19 and then that is reflected in the remote (your fork of aports), but you can't really fix that in the webui to my knowledge 2021-10-12 17:50:42 the git stash/pop should let you just put your single commit of catfish changes into a new branch based on master, then you push that branch and merge request it 2021-10-12 17:51:41 Alright 2021-10-12 17:51:42 I need to read up way more 2021-10-12 17:51:43 Thank you 2021-10-12 17:52:36 you can also, git checkout master, git checkout -b catfish, git cherry-pick $commit-id-of-catfish-commit, if you really don't care about the branch you were just in and fixing it for something else 2021-10-12 17:52:49 but just play around with it and have fun 2021-10-12 17:53:12 Fun... Right haha 🤣 2021-10-12 17:53:34 an ounce of learning is an ounce of fun :) 2021-10-12 17:55:06 https://gitlab.alpinelinux.org/Saijin-Naib/aports/-/commits/Saijin-Naib-catfish/ 2021-10-12 17:55:10 Is this proper/clean? 2021-10-12 17:55:21 Deleted my fork, re-fork, and then made the commit via the UI ahha 2021-10-12 17:57:52 that is private i think 2021-10-12 17:58:09 and sure, that's one way of doing it :) 2021-10-12 17:59:02 oh nvm, i was just logged out randomly 2021-10-12 17:59:14 GitLab seems to do it to me weekly or so 2021-10-12 17:59:14 yep, lgtm 2021-10-12 17:59:24 Even if I have the checkbox for keep my signed in on 😐️ 2021-10-12 17:59:34 Awesome, thanks again, all! 2021-10-12 17:59:59 Subjectively, I think this was much less of a mess. Maybe less heartburn on alpine-devel when it hits 🤣 2021-10-12 18:00:52 i don't think you need the :: rename in source because the package is named the same as what you're renaming it to 2021-10-12 18:01:15 oh nvm, i missed the linebreak 2021-10-12 18:01:50 the uh, archive version posted above at https://archive.xfce.org/src/apps/catfish/4.16/ seems to have properly named sources anyway 2021-10-12 18:01:56 should use that instead 2021-10-12 18:02:35 crap 2021-10-12 18:02:41 Gotta nuke and refork again haha 2021-10-12 18:02:58 not at all 2021-10-12 18:03:10 just undo the commit with what i showed, make the change, then force push it 2021-10-12 18:03:36 you can now clone this branch to yourself after making it online and work with it 2021-10-12 18:03:40 so clone master branch down, or clone my working branch down? 2021-10-12 18:03:55 clone the repo, check out the working branch 2021-10-12 18:04:45 Do I need to remove my MR? 2021-10-12 18:04:54 what is the existing mr 2021-10-12 18:05:02 ah i see it 2021-10-12 18:05:04 uhh, no 2021-10-12 18:06:02 I think I opened the MR against myself. Damnit haha 2021-10-12 18:06:31 yep 2021-10-12 18:06:35 somehow i missed that x) 2021-10-12 18:06:37 then yes 2021-10-12 18:09:13 If I'm grabbing the source from the XFCE archive, should I change the URL to their homepage for catfish, or keep the GitLab page for catfish's source? 2021-10-12 18:10:17 url= should be the homepage 2021-10-12 18:17:12 https://docs.xfce.org/apps/catfish/start 2021-10-12 18:17:38 looks like the right place at a glance 2021-10-12 18:18:00 So in the future, project homepage if it exists, not sourcecode repository 2021-10-12 18:18:12 pretty much 2021-10-12 18:18:24 i don't think anyone would really check either way 2021-10-12 18:19:46 Fair, but I'm trying not to rock the boat any more than I already am. If there are conventions or "understood" things in the community I want to try and catch them where I can 2021-10-12 18:20:10 you ain't rockin anything :) 2021-10-12 18:20:19 everything is okay 2021-10-12 18:20:47 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/WRQEEXssRqWFLpIYaTTHeaOP 2021-10-12 18:21:07 What do I do with this? It built fine without these in the makedepends. 2021-10-12 18:21:26 Check on hardware in XFCE before I push the APKBUILD to make sure it actually works? 2021-10-12 18:22:33 you should always check 2021-10-12 18:22:42 Alright 2021-10-12 18:22:48 build passing is not a check 2021-10-12 18:22:51 Fair 2021-10-12 18:22:54 do you by any chance have like gtk+-dev installed 2021-10-12 18:23:25 or er gtk+3.0-dev 2021-10-12 18:23:38 mmm 2021-10-12 18:24:08 [~/Sources]$ apk list --installed | grep gtk 2021-10-12 18:24:10 Nope 2021-10-12 18:24:46 ah i see 2021-10-12 18:24:50 it needs gtk at runtime 2021-10-12 18:24:52 137 packages right now. Trying to keep it as bare-metal as possible so I can build without missing packages by having them on host 2021-10-12 18:25:14 Alright, awesome 2021-10-12 18:25:40 and maybe the py3 package for gtk or w/e 2021-10-12 18:25:58 you'll figure it out 2021-10-12 18:26:13 Maybe, yeah, haha 2021-10-12 18:26:17 Thanks for the vote of confidence 2021-10-12 18:31:31 py3-gobject3 gdk-pixbuf pango glib mlocate py3-pexpect gtk+3.0, should be most of them 2021-10-12 18:31:56 for depends= 2021-10-12 18:37:50 Hmm, alright. Thank you! 2021-10-12 18:38:01 That's from their list, right, just package names adjusted for how Alpine has them? 2021-10-12 18:38:46 yep 2021-10-12 18:41:50 Is there anything special required to compile GCC? Its failing with `No such package: .makedepends-gcc` after an abuild -r 2021-10-12 18:43:15 amk: that's a symptom 2021-10-12 18:43:44 It means it failed to install dependencies earlier 2021-10-12 18:44:41 ah yes, something is weird other packages are also failing to build 2021-10-12 18:47:05 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/dOThFIXZlNVWVUjfFajJsPnu 2021-10-12 18:47:10 tried adding musl-locales, but did not address it 2021-10-12 18:47:21 Something not built in our py3? 2021-10-12 18:47:34 ah i see, i needed to do an apk fix 2021-10-12 18:48:43 https://stackoverflow.com/questions/7514071/localetextdomain-fails-when-lang-is-not-initially-set 2021-10-12 18:48:59 echo $LANG 2021-10-12 18:48:59 C.UTF-8 2021-10-12 18:49:19 Wrong language/locale set by default and catfish can't handle it? 2021-10-12 18:49:41 musl-locales 2021-10-12 18:49:48 try adding that 2021-10-12 18:49:51 I added that 2021-10-12 18:50:11 Do I need to exec it somehow? 2021-10-12 18:50:14 and maybe musl-locales-lang 2021-10-12 18:50:16 A post-install script? 2021-10-12 18:50:19 nope 2021-10-12 18:50:41 the error isn't the locale error from what you posted 2021-10-12 18:51:18 Oh, the stackexchange thing isn't relevant? 2021-10-12 18:54:48 it's from locale.textdomain being missing, which just means either some package is missing to have it or it will always be missing 2021-10-12 18:54:52 idk where to actually look for it 2021-10-12 18:59:04 https://docs.python.org/3.9/library/locale.html 2021-10-12 18:59:09 Supposed to be part of Python3 2021-10-12 19:01:42 yes, but locale stuff is dependent on libc, it's probably missing due to some musl thing 2021-10-12 19:01:46 sadly i can't help you here 2021-10-12 19:02:29 Ah, poop 2021-10-12 19:02:34 Alright, catfish is dead in the water 2021-10-12 19:02:51 Probably why Nataneal didn't add it to the rest of the XFCE4 stuff he provides 2021-10-12 19:04:21 i am sure someone knows more about it, you could ask in devel too 2021-10-12 19:04:29 gettext on alpine is... complicated 2021-10-12 19:05:16 Complicated is not good for me right now, haha 2021-10-12 19:05:22 I'm hoping for simple 2021-10-12 19:05:42 in this case, "Python applications should normally find no need to invoke these functions, and should use gettext instead." 2021-10-12 19:05:45 It looks like other folks have hit this on StackExchange, and I guess the musl-locales package doesn't provide everything 2021-10-12 19:06:10 Does the "except when they link to other C libraries" part apply? 2021-10-12 19:06:11 odds are they actually mean gettext.textdomain 2021-10-12 19:06:56 in this case they are unconditionally calling locale.textdomain('catfish') 2021-10-12 19:07:30 so the only case where it would apply is if gtk needed to access catfish strings directly, which seems highly unlikely 2021-10-12 19:11:18 Hm... Possible bug for upstream to consider? 2021-10-12 19:31:43 Saijin_Naib[m]: Yeah, I would open a bug report upsteam and try to figure out a patch that will fix it. 2021-10-12 19:56:30 Alright, done. Thanks, all! 2021-10-12 19:56:42 On to the next thing, I guess 2021-10-12 19:56:46 :) 2021-10-12 20:05:00 How bad do you think gufw will be? hahaha 2021-10-12 20:09:12 i'm not sure gobject for webkit exists 2021-10-12 20:09:25 or if that specific gnome icon theme does 2021-10-12 20:09:35 mmm, fair 2021-10-12 20:09:40 aside from that, probably not that bad 2021-10-12 20:09:47 Ideally, it'd be a GUI for awall since Alpine has its own thing 2021-10-12 20:12:52 I'm trying to run `abuild rootbld` in an Alpine chroot on Ubuntu and getting a bwrap permissions error: "bwrap: No permissions to creating new namespace, likely because the kernel does not allow non-privileged user namespaces. On e.g. debian this can be enabled with 'sysctl kernel.unprivileged_userns_clone=1'." 2021-10-12 20:13:12 But my kernel already has that sysctl enabled. Any other ideas? 2021-10-12 20:16:24 user namespaces are blocked in chroot 2021-10-12 20:16:50 yes, the error message is very misleading 2021-10-12 20:17:19 you can replace chroot with bwrap 2021-10-12 20:25:38 Ah, ok. Thanks. 2021-10-12 20:37:10 Is anyone else using XFCE4 as their DE? 2021-10-12 21:01:33 Saijin_Naib[m] - I am 2021-10-12 21:08:56 Ah, awesome. Do you have thumbnails for RAW files, OpenDocument files, videos, fonts, audio files, ePUB documents, etc? 2021-10-12 21:09:35 I'm trying to figure out if the Alpine tumbler is built without most of those things, or if the support libraries are just not installed 2021-10-12 21:09:48 And if it is being built without them, how to make it so they're included 2021-10-12 21:10:25 https://gitlab.xfce.org/xfce/tumbler/-/tree/master/plugins 2021-10-12 21:11:00 do you have libopenraw for the first 2021-10-12 21:11:14 I do, yep 2021-10-12 21:11:34 i do not 2021-10-12 21:11:59 libraw 2021-10-12 21:13:20 So I need to get libopenraw ported first 2021-10-12 21:14:00 looks like it 2021-10-12 21:14:28 https://pkgs.alpinelinux.org/contents?branch=edge&name=tumbler&arch=x86_64&repo=community 2021-10-12 21:14:53 Still seems like a few are missing here compared to upstream 2021-10-12 21:15:00 Is that because we're missing support libraries? 2021-10-12 21:15:09 They're disabled at build time? 2021-10-12 21:16:55 they're not enabled or disabled, the configure automatically adds those cause they're in makedepends 2021-10-12 21:17:08 if you port libopenraw and throw it in it should then get built and thrown in too 2021-10-12 21:17:18 probably 2021-10-12 21:17:29 ditto for whatever random missing thing i guess 2021-10-12 21:45:37 Sweet 2021-10-12 21:45:47 Setting up the APKBuild is so hard haha 2021-10-12 21:46:03 source="https://$pkgname.freedesktop.org/download/$pkgname-$pkgver.tar.bz2" 2021-10-12 21:46:03 builddir="$srcdir/$pkgname-$pkgver/" 2021-10-12 21:46:15 [~/Sources/aports/testing/libopenraw/src]$ ls 2021-10-12 21:46:15 libopenraw-0.3.0 libopenraw-0.3.0.tar.bz2 2021-10-12 21:46:15 after you've done a few, it gets easier\ 2021-10-12 21:46:23 What super-obvious thing am I missing now?" 2021-10-12 21:46:49 Do i need to escape things in the builddir? 2021-10-12 21:47:02 Do you get an error? 2021-10-12 21:47:20 btw, that's the default builddir definition 2021-10-12 21:47:41 https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/abuild.in#L2784 2021-10-12 21:47:43 Whenever I do a newapkbuild it doesn't have that in my APKBuild 2021-10-12 21:47:54 default means it's implicit without being provided 2021-10-12 21:48:14 It just has builddir="$srcdir/" 2021-10-12 21:48:32 yeah cause usually you have to fill it out 2021-10-12 21:48:32 Ahh 2021-10-12 21:48:36 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/MjsKdxRYZUDdrBcbkwYibZKh 2021-10-12 21:48:39 if it's the default you can delete it tho 2021-10-12 21:49:09 sounds like you have $pkgdir in there? 2021-10-12 21:49:41 libopenraw/pkg/libopenraw is $pkgdir 2021-10-12 21:49:57 But you can just leave the builddir out in this case 2021-10-12 21:53:38 Ah, awesome, thanks. That helped a ton 2021-10-12 22:05:48 Build kicks off fine. I think I have a build step configured wrong, though 2021-10-12 22:05:52 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/hKRCAcfHjxXfmBcVvIrnPhIF 2021-10-12 22:06:09 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/GTQHbgdoMuaqVoJOorKhCkPk 2021-10-12 22:07:31 probably should be `./configure --prefix="$pkgdir"/usr` ? 2021-10-12 22:08:32 hmm 2021-10-12 22:08:36 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/VSxBhVnsLtflbInmfpQbHnvb 2021-10-12 22:08:54 I'm referencing this example I was sent prior: 2021-10-12 22:08:54 https://git.alpinelinux.org/aports/tree/community/cmus/APKBUILD 2021-10-12 22:11:42 Sweet, all the way through 2021-10-13 01:36:51 How would one go about testing a library package like libopenraw? 2021-10-13 01:37:01 We obviously don't have anything that depends upon it... 2021-10-13 01:37:20 Try to recompile tumblr and see if it adds the RAW image thumbnailer service during configure? 2021-10-13 01:37:32 tumbler. Blegh. Too much internet 2021-10-13 02:01:20 Not sure if this is part of the scope of the virt kernel, but it doesn't have the uinput module enabled so spice-vdagent doesn't work. 2021-10-13 02:06:21 (which is used in desktop virtual machines for e.g. clipboard sharing) 2021-10-13 02:09:58 good morning 2021-10-13 02:10:52 I was wondering, do we package xcursor themes somewhere? My cursor in bright light is a bit small and I was checking out https://wiki.archlinux.org/title/Cursor_themes, but did not locate the right package on alpine nor the right path in which the xcursor themes would be stored 2021-10-13 03:33:08 @jeffreybenzos I saw on the IRC log that you started to make a suggestion for my failed update due to the power failure. I am running the stable release 2021-10-13 03:34:55 ktprograms: i think quite few people use qemu-ga/vdagent 2021-10-13 03:35:29 you can open an issue (preferably with MR), i think uinput sounds ok as long as it doesn't come with too many deps 2021-10-13 03:35:49 seems like evdev is already enabled so probably not much 2021-10-13 03:37:33 Ok will do 2021-10-13 03:41:49 Maybe I didn't tag the nick right... 2021-10-13 03:43:17 @jeffreybenzos[m] I saw your message on the IRC logs from yesterday regarding me loosing power during an upgrade. I am running the stable release. 2021-10-13 04:08:40 What does lxappearance do to change mouse cursor theme that I can't do manually? I've tried adding a /usr/share/icons/default/index.theme that inherits the cursor theme. 2021-10-13 05:09:57 My filesystem is booting readonly after a failed upgrade 2021-10-13 05:10:48 the-count: are you in an emergency shell? 2021-10-13 05:18:37 I'm not in an emergency shell 2021-10-13 05:20:20 I just recalled that I had the testing repos enabled when I was updating. I'm currently booted into the alpine extended iso. 2021-10-13 05:26:25 just having the testing repos enabled should not break your system 2021-10-13 05:27:23 The main issue was that my battery died in the middle up upgrading 2021-10-13 05:27:47 ok 2021-10-13 05:28:13 so you're on a liveiso now, can you mount your filesystem? 2021-10-13 05:28:31 Yes, it is already mounted 2021-10-13 05:28:34 ok 2021-10-13 05:28:46 also mount your boot system on $mountpoint/boot 2021-10-13 05:28:57 boot partition 2021-10-13 05:30:13 ok, give me a minute 2021-10-13 05:30:56 got it mounted 2021-10-13 05:31:44 then mount --rbind /proc $mountpoint/proc 2021-10-13 05:31:52 same for /sys and /dev 2021-10-13 05:35:43 Ok, all three should be mounted 2021-10-13 05:35:59 ok, now chroot $mountpoint /bin/sh 2021-10-13 05:36:18 ok 2021-10-13 05:36:26 apk fix 2021-10-13 05:37:52 Any idea why fdisk -l shows that my /dev/vda has 41943040 sectors but last usable sector is 20971486? (It's a VM and I've increased the size of the virtual disk) 2021-10-13 05:39:14 cgdisk says the secondary header's self-pointer doesn't reside at the end of the disk. Is that related? 2021-10-13 05:40:25 ikke: Says gtreamer installed and nothing else 2021-10-13 05:40:59 the-count: apk audit --system 2021-10-13 05:41:37 Oh, nevermind, I did 'e' (relocate backup data structures to end of disk) and now it's fine. 2021-10-13 05:42:12 ikke: that completed 2021-10-13 05:44:27 Why is it that disks partitioned with the alpine installer causes fdisk/gdisk to say there's a gap between the main partition table and first usable sector? Is there anything that should be changed in the installer and is it ok to do as it suggests and move the main partition table? 2021-10-13 05:46:34 ikke: I a running apk upgrade and ir seems to be updating fine... Hopefully there wasn't something else required first 2021-10-13 05:47:12 wow, I can't type 2021-10-13 05:47:18 the-count: I don't think so 2021-10-13 05:48:53 now if I disable the testing repos, will it apk downgrade packages or update them when newer versions get moved to main? 2021-10-13 05:50:54 the-count: packages in testing are not in main 2021-10-13 05:50:59 a package is just in a single repository 2021-10-13 05:52:03 oh, right, i should have known 2021-10-13 05:56:43 On some of the scrips it is running post update, it can't create /dev/null. Should I mount that on my filesystem mountpoint as well? 2021-10-13 05:57:37 hmm, that should've come with mount --rbind /dev $mountpoint/dev 2021-10-13 06:01:51 dev has folders inside, could that be it? 2021-10-13 06:02:50 my mistake, I mounted sys onto dev 2021-10-13 06:17:39 What is /sbin/mkmntdirs for? 2021-10-13 06:22:24 ikke: Thank you for your advice! There are a few more issues to iron out but the system is booting normally now! 2021-10-13 06:32:35 ikke: Thank you for your advice! There are a few more issues to iron out but the system is booting normally now! 2021-10-13 06:56:35 hello 2021-10-13 06:56:46 i run into kernel panic for my newly installed alpinelinux on my VM 2021-10-13 07:10:26 using alpine-standard-3.14.2-x86_64 on virtualbox 2021-10-13 07:11:03 Guest2671: What does the panic message say? 2021-10-13 07:11:38 ktprograms: kernel panic not syncing 2021-10-13 07:13:16 Guest2671: Look further either back or forward. There is normally a more informative error message 2021-10-13 07:13:42 problem with the storage medium? 2021-10-13 07:13:52 Guest2671: have you tried using virt iso? 2021-10-13 07:14:39 clandmeter, i can give them a try. i switch to an older build of alpinelinux alpine-standard-3.13.6-x86_64 . see if this help 2021-10-13 07:46:29 is there a good option for searching/duckduckgo/bravesearch etc from cli? 2021-10-13 07:49:18 frag: There is surfraw, but that is just shorthand for opening browser. The only search (meta)engine I know that supports parseable results is searx 2021-10-13 07:50:14 I used to have scripts to pipe searx results into fuzzy selector (rofi) if you want inspiration, but nothing too polished. 2021-10-13 07:52:17 hmm weird there isnt something like that, i found codemadness-frontends in repo, works for youtube (didnt test reddit and ddg didnt seem to work) 2021-10-13 07:52:32 i guess it would have to be kinda generic to be useful enough 2021-10-13 07:52:35 frag: ddg not working? 2021-10-13 07:52:44 ddg let you search on cli 2021-10-13 07:52:49 unless it is broken 2021-10-13 07:52:51 Guest2671: not from codemadness-frontends seems 2021-10-13 07:53:05 Guest2671: in what way? 2021-10-13 07:55:50 ah you can use 'wget "https://html.duckduckgo.com/html/?q=test"' at least 2021-10-13 07:59:46 frag: You could take a look at ddgr (I haven't tested it myself but I found it in the repo) 2021-10-13 08:02:58 ktprograms: ah yes in edge :] 2021-10-13 08:03:13 prob easy to install from git 2021-10-13 08:03:41 frag: Actually what's wrong with edge? Unstable? 2021-10-13 08:04:05 potentially i guess 2021-10-13 08:04:22 frag: ok thanks 2021-10-13 08:06:15 edge is mostly stable, just some small problems from time to time 2021-10-13 08:08:16 So is edge/testing even less stable? 2021-10-13 08:24:28 yes 2021-10-13 08:57:01 the-count: yeah, did you pull stuff over from the live disk? should be able to hack together what you need from the install scripts themselves (setup-alpine, etc) 2021-10-13 11:14:21 Is there any way for apk to install the makedepends for a package (not using abuild)? 2021-10-13 11:16:42 apk does not know about makedepends 2021-10-13 11:18:04 ikke: How come when building packages using abuild -r it shows it's installing .makedepends-$pkgname ? Is that only during build time? 2021-10-13 11:18:17 Yes 2021-10-13 11:18:24 Virtual packages 2021-10-13 11:19:18 ikke: Ok, thanks 2021-10-13 11:19:29 apk add -'t .makedepends-foo pkg1 pkg2 ... 2021-10-13 11:21:04 ikke: Thanks 2021-10-13 11:21:19 -' 2021-10-13 13:16:32 Hi. If I remove a package with apk del , will dependencies also be removed 2021-10-13 13:16:35 ? 2021-10-13 13:17:01 Yes 2021-10-13 13:17:19 As long as they are not required by other packages 2021-10-13 13:17:51 Ok thanks 2021-10-13 13:18:31 How do I list installed packes? 2021-10-13 13:18:49 List -I 2021-10-13 13:18:53 :) 2021-10-13 13:19:11 can take a pattern 2021-10-13 13:23:04 Fit on one screen so it's ok :) 2021-10-13 13:25:30 Hi. I would like to avoid switching into framebuffer mode at boot time. Where can I disable that? 2021-10-13 13:26:48 Just a quicky. I'm setting up a dedicated alpine instance for mariadb. Anything specific I should think of? 2021-10-13 13:27:37 the usual system specs for your database, nothing alpine specific immediately springs to mind 2021-10-13 13:46:25 lopid: thanks 2021-10-13 13:50:10 TRA: add `nofb` to the cmdline? 2021-10-13 13:50:50 depends on whether it's grub or syslinux or something else 2021-10-13 14:01:46 i have a qemu running for making my apk test builds. the biggest stuff thats on the virtual harddrive are the firmware files. i read in the wiki that i should just install the ones needed... but whats the best way to find them out? 2021-10-13 14:01:53 and how can i remove firmware packages? 2021-10-13 14:01:59 arent there larger dependencies? 2021-10-13 14:03:54 dmesg | grep firmware 2021-10-13 14:04:21 you will see every firmware file loaded 2021-10-13 14:04:23 @psykose this is syslinux - plain x86 boot 2021-10-13 14:04:39 throw them onto the package file search, see what firmware packages are actually needed 2021-10-13 14:04:54 psykose: its empty here 2021-10-13 14:05:14 then just linux-firmware-none 2021-10-13 14:05:16 Remove firmware: apk add linux-firmware-none 2021-10-13 14:05:16 2021-10-13 14:05:19 YEAH :) 2021-10-13 14:05:21 thanks 2021-10-13 14:05:37 TRA: /etc/update-extlinux.conf, the APPEND section 2021-10-13 14:06:06 er 2021-10-13 14:06:09 ok apk add linux-firmware-none doesnt any installed firmware 2021-10-13 14:06:25 default_kernel_opts 2021-10-13 14:07:05 ok apk del linux-firmware after that 2021-10-13 14:07:09 then you want to rerun the update 2021-10-13 14:07:10 @psykose Thanks! 2021-10-13 14:07:39 update-extlinux afterward 2021-10-13 15:49:16 Hey, what's the prefered way of disabling the root user with alpine ? passwd -l root ? (I have doas already) 2021-10-13 15:50:37 disable SSH login, set up a good password and leave it alone? 2021-10-13 15:50:49 see the -l description 2021-10-13 15:51:19 make the root shell /sbin/nologin 2021-10-13 15:51:47 I'd rather not even have a password for it 2021-10-13 15:52:26 then just passwd -d and -l 2021-10-13 15:55:08 thx, I was afraid that doing that, I would lose doas access along with it 2021-10-13 15:56:02 You won't 2021-10-13 15:59:28 i don't think you'll really gain anything by doing it 2021-10-13 16:07:21 Hello, can I install Alpine on qemu-system-arm? 2021-10-13 16:07:52 Yes 2021-10-13 16:40:40 Well, I tried 'virt' machine type, and it didn't work - there were nothing after 'switching to SVC with MMU enabled' or something like this, however, qemu continued to consume cpu. Which machine type should I choose? (Running Alpine Virt) 2021-10-13 17:04:44 When I'm making the APKBUILD, does anyone care about the order I list packages in the various *depends lines? Alphabetical? The order I discover I need them? 2021-10-13 17:05:30 if they get shown by an ordinary user listing deps, alphabetical 2021-10-13 17:05:50 How do I know that? 2021-10-13 17:06:14 aren't you asking in order to make a decision about it? 2021-10-13 17:06:23 Yes 2021-10-13 17:06:36 but you don't know what the deps are? 2021-10-13 17:06:48 Not always, no 2021-10-13 17:06:58 https://gitlab.xfce.org/thunar-plugins/thunar-shares-plugin 2021-10-13 17:07:07 This one has me going in without any idea, for instance 2021-10-13 17:07:22 So the first I discovered was intltool, then gtk+3.0-dev 2021-10-13 17:08:45 unless they can be displayed in some sort of tree hierarchy, then i'd say the order you discover them is unimportant 2021-10-13 17:09:26 Okay 2021-10-13 17:09:43 So putting them into the APKBUILD I'm submitting in discovered order is okay 2021-10-13 17:09:44 Thanks 2021-10-13 17:09:51 it looks like apk might short them alphabetically anyway 2021-10-13 17:09:57 s/short/show 2021-10-13 17:10:42 libs, at least 2021-10-13 17:11:16 Saijin_Naib[m]: Some packages use some kind of logical order, but if the amount of dependencies becomes larger, people tend to put one dependency per line and sort them 2021-10-13 17:12:12 Okay, that makes sense. I assume something like the APBUILD for Firefox would be an example of that I can reference? 2021-10-13 17:12:29 I'm hoping that these next few APKBUILDs will be simple. har. 2021-10-13 18:18:01 https://gitlab.freedesktop.org/libopenraw/libopenraw 2021-10-13 18:18:15 When a licenese is listed as GNU LGPLv3 2021-10-13 18:18:20 Is that -only or -or-later 2021-10-13 18:18:30 -only, right? 2021-10-13 18:18:40 Since we can't assume they'll accept further revisions of the license? 2021-10-13 18:18:44 Saijin_Naib[m]: It should be mentioned in a source header 2021-10-13 18:19:55 https://gitlab.freedesktop.org/libopenraw/libopenraw/-/blob/master/lib/bititerator.cpp#L10 2021-10-13 18:20:08 https://gitlab.freedesktop.org/libopenraw/libopenraw/-/blob/master/include/libopenraw/libopenraw.h 2021-10-13 18:20:21 Ah, okay 2021-10-13 18:20:25 Thank you 2021-10-13 18:38:06 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/26398/diffs#note_185270 2021-10-13 18:38:16 Can anyone help me with this review? 2021-10-13 18:38:40 can I just add the --build=$CBUILD --host=$CHOST after the --prefix=/usr 2021-10-13 18:38:43 you just have to add the line 2021-10-13 18:39:27 So the correct line would be:... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/ViVPuWdDnjliYxqxbrKqfkYm) 2021-10-13 18:39:29 ? 2021-10-13 18:39:51 Does this apply for all "compiled" packages that aren't just python? 2021-10-13 18:42:04 yep 2021-10-13 18:42:52 Awesome. Thanks. 2021-10-13 18:48:12 Is it normal for this to fail now with the --build and --host options in place? 2021-10-13 18:49:09 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/lrqsBSBeuciJoalldlFBkcNz 2021-10-13 18:49:26 I'm building in WSL2 2021-10-13 18:49:30 If that matters 2021-10-13 18:49:53 add update_config_sub to prepare() 2021-10-13 18:50:05 some packages have old versions of these scripts 2021-10-13 18:51:23 Oh wow, that was it. Thank you 2021-10-13 18:54:18 thank you lord ikke 2021-10-13 18:56:16 Seriously. That's some good stuff 2021-10-13 18:56:49 https://gitlab.xfce.org/bindings/thunarx-python 2021-10-13 18:57:01 This has .c and .h files, so this needs the --build and --host options, right? 2021-10-13 19:03:35 yep 2021-10-13 19:03:56 to be precise you need those as arguments to ./configure, some packages might not have a configure 2021-10-13 19:04:48 Okay, that makes sense 2021-10-13 19:05:04 Now the update_config_sub step in prepare, is it normal for that to fail if it isn't needed? 2021-10-13 19:05:19 I added that to my thunarx-python APKBUILD and it causes the build to abort after saying it isn't needed 2021-10-13 19:05:29 Should that not fail gracefully and continue? 2021-10-13 19:05:31 sounds right, idk 2021-10-13 19:05:34 I remove it and it builds fine 2021-10-13 19:05:44 Hmm, okay 2021-10-13 19:07:46 the comment for !check means to put a #comment in the apkbuild for why 2021-10-13 19:08:20 Yeah. I had it disabled because they had a warning in their GitLab that concerned me 2021-10-13 19:08:32 I'm not sure it applies to the normal test though, so I've enabled the standard test 2021-10-13 19:11:39 i think you misread it, the default doesn't download anything 2021-10-13 19:11:55 cd testsuite; ./testsuite does 2021-10-13 19:12:02 and with !check your defined check doesn't run 2021-10-13 19:14:27 and make check needs that bootstrap of testsuite to run first as far as i understand it, so you should delete your check function 2021-10-13 19:14:35 and leave the !check 2021-10-13 19:14:44 unless it runs differently than i guess :) 2021-10-13 19:15:24 Oh 2021-10-13 19:15:31 it runs tests 2021-10-13 19:15:49 Let me capture some log 2021-10-13 19:16:16 sure 2021-10-13 19:16:47 I'm hoping the test suite knows what to do on its own haha 2021-10-13 19:16:59 Unless these tests are all passing since they're not testing anything at all 2021-10-13 19:19:00 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/dRUKRJfOQNQFtxJyBXeucTtX 2021-10-13 19:19:15 they all passed 2021-10-13 19:19:23 looks like they ran to me 2021-10-13 19:19:26 Yeah. Is that weird if I'm not supplying RAWs? 2021-10-13 19:19:39 i don't have the slightest clue 2021-10-13 19:19:39 I have NFI if this is even testing what it is supposed to at this point, haha 2021-10-13 19:19:44 but it's running something 2021-10-13 19:19:52 so remove the !check and all is good 2021-10-13 19:19:59 usually software has own test data 2021-10-13 19:19:59 I'm hoping it is testing funcitonality that does not require sample RAWs. 2021-10-13 19:20:20 They say they don't provide any unless you bootstrap, and that grabs random stuff without validating them apparently 2021-10-13 19:20:21 or it feeds arbitrary values 2021-10-13 19:20:34 No, an automated test suite wouldn't require the user to provide data. 2021-10-13 19:20:34 i mean these tests have names 2021-10-13 19:20:34 Oh, that makes sense 2021-10-13 19:20:40 and none of the names are about raws 2021-10-13 19:20:46 just the other parts 2021-10-13 19:20:50 Yeah. 2021-10-13 19:20:53 which is good enough 2021-10-13 19:21:56 Alright, so my 3rd (meep) commit has the test suite enabled and I think all of Leo's reviews addressed 2021-10-13 19:22:45 If so, I have to see if it can be added to the makedepends for Thunar so it can RAW thumbnail support 2021-10-13 19:22:55 Do I just open an issue for that to discuss first? 2021-10-13 19:23:23 You could straight away create an MR 2021-10-13 21:21:06 https://gitlab.alpinelinux.org/Saijin-Naib/aports/-/pipelines/96523 2021-10-13 21:21:31 Does the failure under build-s390x mean I should drop it from the arch= for my APKBUILD? 2021-10-13 21:21:51 yes 2021-10-13 21:21:52 (assuming I lack the capability to fix/patch/port whatever is needed) 2021-10-13 21:21:59 Okay, I'll do that then 2021-10-13 21:22:12 I recommend looking at other rust packages 2021-10-13 21:22:28 Does rust not work under that arch? 2021-10-13 21:22:29 `# s390x, mips64 and riscv64 blocked by cargo` 2021-10-13 21:22:33 This one isn't rust I don't think. 2021-10-13 21:22:46 at minimum, you should have this `arch="all !s390x !mips64 !riscv64"` 2021-10-13 21:22:50 it's mostly not rust, but it has rust as a dependency 2021-10-13 21:22:59 for one part of it 2021-10-13 21:23:04 Ahh, yep. 2021-10-13 21:23:06 hence, yeah 2021-10-13 21:24:12 Thanks again for the save, haha 2021-10-13 22:59:24 anyone got tesseract from repo to work? it just seems to work and work forever, trying: tesseract /4/photo.jpg test.txt 2021-10-14 01:55:20 does anyone know how to have root on zfs on a pi4? I am struggling with this 2021-10-14 01:59:05 I'm not sure if the packages for creating that FS are included in the boot media. 2021-10-14 02:00:09 Is there any way to check which packages I have from which repo? 2021-10-14 02:01:18 As in I need to figure out why I have edge/testing enabled. I can't remember which package I enabled it for. 2021-10-14 02:01:57 I inistalled zfs, but when I do modprobe I get a module not found error 2021-10-14 02:02:41 isn't the zfs module through dkms/akms 2021-10-14 02:02:43 It is possible that it isn't compiled as a kernel module for the Pi images 2021-10-14 02:03:56 ah, apparently on extended has it 2021-10-14 02:04:05 if I install the os on the card with ext4, I do get the module. 2021-10-14 02:04:49 only the extended image starts with it i guess 2021-10-14 02:08:05 would it be possible for me to install using ext4 then install the zfs-rpi4 and rebuild the file system? 2021-10-14 02:09:35 How do you have your bootable image written? Editable FS, or dd-like? 2021-10-14 02:09:35 Maybe grab the apks and put it in the drive so you can add it during install? 2021-10-14 02:09:52 I don't know if that'd work... Hmm 2021-10-14 02:09:59 Chroot in and add it? 2021-10-14 02:10:41 you could install onto the sd from your computer 2021-10-14 02:10:45 the issue is I can't find the module to modprob it 2021-10-14 03:36:20 When networking is starting I get ip: ioctl 0x8913 failed no such device udhcpc: 0x8933 failed: no such device ip: can't find device 'wlan0' 2021-10-14 03:39:34 if you run `ip a` is there a wlan0 2021-10-14 05:27:22 Anyone have the spare energy to help me un-mess what I'm doing with my MR 2021-10-14 05:27:32 I was asked to squash my commits, and that has now become a whole thing 2021-10-14 05:29:26 Saijin_Naib[m]: what's the problem? 2021-10-14 05:29:50 Fundamentally, none of this is what I do so I'm trying to piece things otgether from StackExchange and messages here 2021-10-14 05:30:10 In an attempt to squash my commits, I ended up reverting an important change and just adding more commits to the MR 2021-10-14 05:30:33 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/26398 2021-10-14 05:30:55 Commit 9a1625a5 should be fully correct 2021-10-14 05:33:07 What I read on SE was this:... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/RiBGWkvQycfhhrpaiAPfDZKv) 2021-10-14 05:33:22 So now you need to make your branch only have that 1 latest commit? 2021-10-14 05:33:26 Yeah 2021-10-14 05:34:01 When I tried to do the above, it wouldn't let me push until I did a git pull, which seemed to have screwed things and just added more commits 2021-10-14 05:34:45 The GitLab WebUI doesn't let you squash or revert changes, so that sucks 2021-10-14 05:51:34 Saijin_Naib[m]: for a start can you please send the output of git log (as a code block or whatever it's called) 2021-10-14 05:59:28 Why does the setup-alpine script not call setup-ntp if it's running in a QEMU virtual machine? 2021-10-14 07:19:19 Hello, why is there no alpine 3.13 virt image for armv7 available? 2021-10-14 08:52:55 I think something broke my ssh setup... manager or something? I have to enter passphrase for key every time I try to ssh to machine 2021-10-14 09:01:43 passphrase for the key implies you don't have it in an agent 2021-10-14 09:02:00 suggests 2021-10-14 09:02:34 Yeah thank you am figuring right now why not 2021-10-14 09:02:44 Might be migrating to fish shell as default shell 2021-10-14 11:30:31 Does ash have a way to sudo the previous command? (with bash it is sudo !!) 2021-10-14 11:34:30 Not from what I know of. You can always use 'sudo ', but that's not as nice to do every time 2021-10-14 11:40:37 yyp: Especially as I don't have a key 2021-10-14 11:40:48 Then you can use vi mode :P 2021-10-14 12:00:01 ^a also gets you to the start of the line 2021-10-14 12:34:48 ikke: Thanks, that works. 2021-10-14 13:08:47 Hello. at AlpineWSL we not have a openrc and crond.. what crond select ? [in|f|d]cron package? 2021-10-14 13:14:40 looks like they're all available 2021-10-14 13:20:14 yes, I can install one and all of them but if I know what will best as default then it's easier to start 2021-10-14 13:31:11 that might be subjective. i'm using busybox crond 2021-10-14 13:32:10 ktprograms: Sorry, my scrollback buffer got cleared apparently -_-. It also looks like Leo saved my MR by rebasing/squashing, I think. 2021-10-14 13:40:54 Saijin_Naib[m]: Great that your MR got through! Basically what you needed to do is interactively rebase from the correct commit (change everything after the first line to 'squash'), fix the conflicts, and then you need to force push (--force or +{branch_name}). Also when the rebasing is done, edit the commit message instead of leaving it as defaults. 2021-10-14 13:41:38 So the snippet I got from StackExchange was mostly the correct flow? 2021-10-14 13:41:54 git pull https://gitlab.alpinelinux.org/Saijin-Naib/aports Saijin-Naib-libopenraw 2021-10-14 13:41:54 git rebase -i Saijin-Naib-libopenraw~4 Saijin-Naib-libopenraw 2021-10-14 13:42:23 You'd need to look in the git log to see how far back you need to rebase 2021-10-14 13:42:36 At the time, the ~4 level was just my commits 2021-10-14 13:42:45 `git rebase -i HEAD~20`, fixup all commits that are unnecessary to be in log 2021-10-14 13:42:56 Some other folks recommended using fixup instead. Does that make sense? 2021-10-14 13:43:12 Saijin_Naib[m]: The force push is needed since squashing changes history. 2021-10-14 13:43:51 Ah, that I did not do. That might have been why it complained and asked me to git pull before pushing 2021-10-14 13:44:12 Yes, but you need to be careful so you don't accidentally delete too much history 2021-10-14 13:44:28 But if your (local) git log looks fine, then push with the --force 2021-10-14 13:45:03 Also I could see from your branch history that after the rebasing you didn't edit the commit message. 2021-10-14 13:45:43 The one that said Merge? 2021-10-14 13:46:00 I wasn't expecting that one to go up. I thought that'd get consumed in the squash 2021-10-14 13:46:38 lopid: maybe I mess stg.. at AlpineWSL I see a crontab -e and /etc/periodic/* files.. but they not work. What means your busybox crond ? I just start check the service status but even openrc was not installed and I think this is main problem, maybe not ? 2021-10-14 13:47:23 Saijin_Naib[m]: The one with about 5 lines of (community/libopenraw: new aport). Do you use vim for the git commit message editor? 2021-10-14 13:49:23 lopid: I see, at clear AlpineWSL I found that `crond --help` works inside busybox, thanks for hint, I will working on this. 2021-10-14 13:49:35 busybox crond is the in-built busybox cron, which uses /etc/periodic files. it's managed just like any other service 2021-10-14 13:50:42 however, if you don't have openrc, managing it or any other service is probably going to be... different. you could try using the /etc/init.d scripts directly 2021-10-14 13:52:08 I tested it .. in AlpineWSL it's like it not start, ps not show it, /etc/init.d is empty folder . This show that AlpineWSL is differ distro then classical Alpine :D 2021-10-14 13:53:10 indeed 2021-10-14 13:53:14 AFAIK, you don't have any init / service system in WSL 2021-10-14 13:54:02 More like docker, though not sure if this changed with WSL2 2021-10-14 13:55:16 No, no init in Alpine WSL2 2021-10-14 13:55:27 I've tried adding/running OpenRC, but it has proven difficult 2021-10-14 13:57:27 openrc can be install but I think about WSL2 I just use external scheduler and workaround by: Alpine.exe run script.sh from Windows Scheduler... WorkARound is like 'clamp band' or WD40 :D 2021-10-14 13:58:20 Thanks for info 2021-10-14 13:58:27 Maybe! I don't understand it very well. You could always ping Hayden Barnes or Craig Lowen, they're quite nice 2021-10-14 14:00:08 run it in a proper vm 2021-10-14 14:00:32 I'd say don't use WSL for such stuff 2021-10-14 14:02:09 AlpineWSL2 is used by me to testing python base differ git hub repo.. better then ubuntuWsl, and I like small distro 2021-10-14 14:12:42 You can still run it as a standard VM though 2021-10-14 16:15:53 does sysctl.d perform the same functionality as sysfs.d? 2021-10-14 16:15:59 Or equivalent-ish 2021-10-14 16:16:34 never heard of sysfs.d 2021-10-14 16:16:47 My backlight driver gives Linux a minimum value of 0, which I suppose is technically correct but 0 for this is actually fully off so the panel is unusable. I want to block that so that 1 is the hard minimum 2021-10-14 16:16:52 I got it from here: 2021-10-14 16:16:52 https://askubuntu.com/questions/1215522/how-to-change-the-minimum-brightness-value 2021-10-14 16:17:06 huh, apparently debian has such a thing 2021-10-14 16:17:15 I know this is Ubuntu and looks like systemd stuff, hence my confusion if we have an analog 2021-10-14 16:17:35 sysfs.d looks like the equivalent for /sys entries on boot 2021-10-14 16:17:38 yeah it's basically the same 2021-10-14 16:17:48 Every time my laptop DPMS the display or sleeps, upon resume/wake it sets backlight to 0 and I'm in trouble 2021-10-14 16:17:51 isn't that just a gnome issue 2021-10-14 16:18:07 i mean the "minimum brightness" 2021-10-14 16:18:21 For me GNOME/GDM weren't re-invoking at all. I'd have to CTRL+ALT+F1 to get it to wake up 2021-10-14 16:18:55 Is that a backlight issue? 2021-10-14 16:18:57 This is different. lightdm/XFCE are alive and at the login screen, I just can't see unless I shine a flashlight just right 2021-10-14 16:19:14 sysfsutils seems to be in packages 2021-10-14 16:20:16 And it seems I pulled it in already 2021-10-14 16:20:18 though there is no service packages that would use it 2021-10-14 16:20:27 But I do not have that directory 2021-10-14 16:20:30 Ah... 2021-10-14 16:20:44 you can enable init.d/local and write a local script 2021-10-14 16:21:26 it runs last, then it runs /etc/local.d/*.start on start and *.stop on stop, you just place an executable shell script in that dir 2021-10-14 16:21:37 could just make it write something to the /sys point you want 2021-10-14 16:21:48 Hmm... 2021-10-14 16:22:01 That sounds like it'd take care of boot/shutdown time, which currently is okay 2021-10-14 16:22:06 Unless I'm misunderstanding 2021-10-14 16:22:43 i am wondering what sets the backlight to 0 on wake though 2021-10-14 16:22:45 i have never seen that 2021-10-14 16:22:48 It'd have to be during runtime or sleep/resume. Whatever is causing the display to go to backlight 0 during usage I need to stop 2021-10-14 16:22:52 it must be a gdm specific thing 2021-10-14 16:22:53 Yeah, exactly. 2021-10-14 16:23:02 Not using GDM though 2021-10-14 16:23:10 lightdm, unless that is related? 2021-10-14 16:23:11 or whatever the login is 2021-10-14 16:23:13 yea 2021-10-14 16:23:40 Could it be busybox-acpi? 2021-10-14 16:23:41 Do I need "real" acpi/acpid? 2021-10-14 16:24:25 I have xfce4-powermanager set to not go below brightness 1, so it shouldn't do it, but when I use the keyboard to set brightness down, it will dip below 1 and turn off the display 2021-10-14 16:25:01 the keyboard shortcuts probably use the default acpi handlers and bypass the xfce4 settings entirely 2021-10-14 16:25:29 though, when it's at 0/off after resume, can't you just use the keyboard to fix that? 2021-10-14 16:25:37 or is it somehow stuck on 0 2021-10-14 16:26:32 It gets stuck until I'm in my session 2021-10-14 16:26:42 hm 2021-10-14 16:26:47 the keyboard shortcuts should work anywhere 2021-10-14 16:26:51 Every display manager I've tried ignores the brightness keys on login/lock 2021-10-14 16:27:06 asdly i can't help much with this 2021-10-14 16:27:11 That's okay 2021-10-14 16:27:14 THank you for trying 2021-10-14 16:27:19 this sort of thing is exactly why i haven't used any display manager in 5 years 2021-10-14 16:27:30 without them things don't get this sort of weird edgecase behaviour 2021-10-14 16:27:37 Wait, really? 2021-10-14 16:27:51 i thought they were supposed to solve a whole bunch of stuff and get the session started nicely 2021-10-14 16:27:55 you can just type `startx` or whatever equivalent to start after logging in on tty 2021-10-14 16:28:03 vs logging into terminal and startx/startxfce4 2021-10-14 16:28:46 i'm not sure what specific extras they do honestly, but i can say i didn't have anything be.. broken, with xfce or i3 back in the x days 2021-10-14 16:28:48 Hah, that's amazing 2021-10-14 16:29:08 Maybe it is crap firmware on this thing. Who knows 2021-10-14 16:29:23 I just barely got the touchpad supported not long ago in Alpine 2021-10-14 16:29:33 my backlight is a bit scuffed without acpi_backlight=native on the cmdline pretty sure, there's always some weird device-specific stuff to do 2021-10-14 16:29:34 I may be missing other things that need modules/drivers and don't knowit 2021-10-14 16:30:41 but if you don't mind the difference in process/aesthetic and want to try just using ole' startx, you could give it a shot 2021-10-14 16:32:22 I am single-user only on here. I thought I needed display manager for lockscreen, which I do want/need since I sometimes leave this laptop where it is accessible 2021-10-14 16:32:30 Unless I'm conflating things and it doesn't provide that 2021-10-14 16:32:49 Hmm, guess it is worth a shot 2021-10-14 16:32:59 Easy as rc-update del lightdm 2021-10-14 16:34:55 for me on sway i can lock then invoke sleep via keyboard and it wakes up locked, i believe that is possible on x but it has really been too long sadly 2021-10-14 16:40:22 All good 2021-10-14 16:40:34 If I may, do you think acpid in Default runlevel is correct? 2021-10-14 16:40:51 Isn't acpi a system-level thing... Is not boot or sysinit better? 2021-10-14 16:41:02 default sounds correct 2021-10-14 16:42:02 acpi is for power management, and some other events or something now on keyboard shortcuts, none of that is needed pre-default 2021-10-14 16:42:10 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/YTyMRnraZCFRTntgtIpUfyxw 2021-10-14 16:42:45 Oh, okay... Guess I'm too used to Windows world where ACPI sort of underpins everything it seems. 2021-10-14 16:42:47 Have you disabled markdown in your client? 🤔 2021-10-14 16:42:53 binfmt should be boot 2021-10-14 16:42:59 hwdrivers should be sysinit 2021-10-14 16:43:50 PureTryOut (matrix.org) At the very least it isn't enabled by default in Fractal. Thank you and sorry 2021-10-14 16:44:03 I'm used to using Element which I think has it on by default 2021-10-14 16:44:16 psykose, thank you. I'll adjust that presently. 2021-10-14 16:44:32 PureTryOut: i notice you are the pipewire maintainer, have you thought about adding wireplumber to packages as well 2021-10-14 16:46:28 Only recently since the old maintainer didn't really use the thing 😛 2021-10-14 16:46:36 hehe 2021-10-14 16:46:42 For wireplumber I'm kinda waiting for upstream to recommend it as replacement for media-session 2021-10-14 16:46:51 aye, makes sense 2021-10-14 16:47:21 the absolute latest release also seemingly has some issue/breaks for me with elogind on gentoo, i think i'll give it another try and see what's up once my new ssd arrives and i put alpine on my laptop 2021-10-14 16:47:58 the only difference over media-session i've noticed is that bluetooth works better for some reason, i haven't really explored much else 2021-10-14 18:34:54 Alpine on a proper SSD is wicked fast. That must be nice! Still fastest I've tested on my eMMC by far 2021-10-14 19:17:00 eMMC can read sata ssd speeds 2021-10-14 19:17:04 reach* 2021-10-14 19:17:16 though perhaps a touch below on average 2021-10-14 19:17:22 and what i ordered is nvme 2021-10-14 19:17:24 should be fun :) 2021-10-14 19:30:11 That I have yet to experience for myself. SATA3 SSDs already feel like they're so fast I couldn't possibly notice anything further, so having 3x+ that bandwidth is... wow 2021-10-14 19:30:36 Yeah, this little cheapo laptop has a decent eMMC, but it is not SATA3-tier 2021-10-14 19:35:36 Far cry from how awful they were not too long ago 2021-10-14 19:35:49 ACTION uploaded an image: (100KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/ujWjDGLGymGokKMJIJPfYoot/image.png > 2021-10-14 19:37:19 Honestly the biggest speed benefit of Alpine is "fits into RAM so you don't really nead to read from disk" 2021-10-14 19:37:55 there's not much of a noticeable difference going from sata to nvme 2021-10-14 19:38:00 just a tad 2021-10-14 19:46:37 if you have enough ram to store the whole system, then you can do that by just find / -xdev -exec cat {} + >/dev/null 2021-10-14 19:47:23 it will be evicted if you run low on ram, but in that case you would otherwise run out of ram 2021-10-15 01:30:08 Why is setup-ntp skipped if running inside QEMU? https://gitlab.alpinelinux.org/alpine/alpine-conf/-/blob/master/setup-alpine.in#L190-192 2021-10-15 01:37:42 vm should be synced to host 2021-10-15 02:02:50 anyone know where the fbcon kernel module is packaged? I'm trying to set up a new machine with an AMD gpu 2021-10-15 02:04:16 (the wiki page for radeon video recommends loading it automatically) 2021-10-15 02:09:18 Hello71: Ok makes sense 2021-10-15 02:09:56 riverdc: it is part of the kernel, and the instructions are obsolete 2021-10-15 02:10:22 kk, I suspected as much 2021-10-15 02:22:24 the rest of that page is also mostly obsolete 2021-10-15 02:23:00 in most cases you are better off reading the arch wiki 2021-10-15 02:23:49 alright, I'll take a look there 2021-10-15 02:24:51 the amdgpu module does need to be loaded, though, correct? that one doesn't seem to have made it into the kernel 2021-10-15 02:36:57 amdgpu was integrated into the kernel more than 6 years ago 2021-10-15 02:40:04 sorry, misspoke. just meant that it doesn't seem to get automatically loaded, whether through /etc/modules or otherwise, unless I do it by hand 2021-10-15 02:40:34 even after installing xf86-video-amdgpu 2021-10-15 02:41:33 your card is not supported by amdgpu 2021-10-15 02:41:56 and xf86-video-amdgpu is for x only, it is unrelated to amdgpu kernel driver 2021-10-15 02:41:59 and also not recommended 2021-10-15 02:42:45 ah, I see 2021-10-15 02:43:08 > your card is not supported by amdgpu 2021-10-15 02:43:31 you're inferring this based on the fact that it wasn't loaded automatically, I guess? 2021-10-15 02:43:52 Did you lspci it or dmesg | grep firmware 2021-10-15 02:46:22 yeah, shows up normally in lspci 2021-10-15 02:47:34 What is it? 2021-10-15 02:47:42 dmesg shows `platform regulatory.0: Direct firmware load for regulatory.db failed with error -2` 2021-10-15 02:47:56 That's wifi, I believe 2021-10-15 02:48:12 There's a regdb package you can load for that 2021-10-15 02:49:04 https://paste.ee/p/stneM for the relevant lspci output 2021-10-15 02:50:34 https://pkgs.alpinelinux.org/package/edge/main/x86_64/wireless-regdb 2021-10-15 02:50:58 thanks 2021-10-15 02:59:57 that is definitely new enough for amdgpu 2021-10-15 03:04:37 indeed. very possible that I've messed something else up. the actual issue I'm having is that sway complains about not being able to open any DRM device, which I assumed was driver/firmware related 2021-10-15 03:04:48 can you post dmesg 2021-10-15 03:05:38 yep give me a moment 2021-10-15 03:07:24 and the sway log 2021-10-15 03:07:35 sway -dV &> log 2021-10-15 03:07:36 or somesuch 2021-10-15 03:15:57 What would one's bugsense say to needing to occasionally boot a Manjaro Linux LiveUSB for hardware brightness keys (FN + F#...) to work in any other Linux OS 2021-10-15 03:16:22 (Alpine on disk, Debian-derived LiveUSB, etc) 2021-10-15 03:16:37 Crap ACPI in system UEFI Firmware? 2021-10-15 03:16:50 https://paste.ee/p/blEMs (sway log) 2021-10-15 03:17:36 What does less /proc/cmdline say? 2021-10-15 03:17:43 https://paste.ee/p/o6TGq (dmesg) 2021-10-15 03:18:35 you do have linux-firmware installed right 2021-10-15 03:18:39 there is no firmware loaded for the gpu 2021-10-15 03:18:49 i do yeah 2021-10-15 03:19:08 every single device that takes firmware normally gets a `firmware` line in dmesg, so since there's nothing something is missing 2021-10-15 03:19:09 this is from a fresh install where I haven't added kms to the mkinitfs features 2021-10-15 03:19:45 and I did a manual modprobe of amdgpu, sway log was identical 2021-10-15 03:20:03 psykose: hmm ok 2021-10-15 03:20:57 Saijin_Naib[m]: i guess you're asking about the modules in cmdline? it's `modules=sd-mod,usb-storage,ext4` and there isn't any nomodeset option 2021-10-15 03:24:49 what firmware does modinfo amdgpu show 2021-10-15 03:25:16 also are you on lts? the firmware is only in edge 2021-10-15 03:25:28 ohhh.... well there's the problem 2021-10-15 03:25:29 it's /lib/firmware/amdgpu/navy_flounder* 2021-10-15 03:25:34 not in the lts firmware 2021-10-15 03:25:53 oh nvm, it is in 3.14 too 2021-10-15 03:25:56 hm 2021-10-15 03:26:15 modinfo should say what's loaded 2021-10-15 03:26:17 i'm using the busybox dmesg btw, not sure if it's any different 2021-10-15 03:27:08 not as far as i know 2021-10-15 03:27:34 yep, amdgpu/navy_flounder_*.bin is there 2021-10-15 03:28:18 you need to use linux-edge for your card 2021-10-15 03:28:59 Saijin_Naib[m]: smells like you are disabling function keys 2021-10-15 03:29:19 right, that makes sense, should have tried that sooner. regardless thanks for the help all, very much appreciated 2021-10-15 03:29:52 Is it ok to use the testing repo while using main? 2021-10-15 03:30:12 Hello71 and booting specifically to Manjaro only fixes it across other boots until the system crashes again? 2021-10-15 03:30:18 each package is in only one repository for each release 2021-10-15 03:30:37 Saijin_Naib[m]: what do you mean "crashes" 2021-10-15 03:30:48 So it's not like having main and edge at the same time which is warned against? 2021-10-15 03:30:55 ktprograms: you can tag it in your repositories file as "@testing" or similar, that has worked well for me 2021-10-15 03:30:57 main is in edge 2021-10-15 03:31:22 sorry, I meant stable 2021-10-15 03:31:30 I think it is advised against to mix packages from edge with a stable like 3.14 2021-10-15 03:31:44 So is stable main and community mixable with edge testing? 2021-10-15 03:31:46 tagging is an antifeature and should be removed from apk. it is consistently used for unsafely combining packages from different releases 2021-10-15 03:31:49 no. 2021-10-15 03:32:05 Hello71: Ok thanks 2021-10-15 03:32:10 if you know what you are doing you can mix it 2021-10-15 03:32:14 since you have to ask, no :p 2021-10-15 03:32:14 alright, listen to Hello71 then, not me 2021-10-15 03:33:14 Hello71 my install will randomly hard lock without warning. Usually after that (and kiss of death power cycle), my brightness FN keys won't work. All others work fine. To remedy this, I simply boot to Manjaro LiveUSB and they work. Boot back to Alpine and all is well until the next unexpected crash. 2021-10-15 03:33:28 Blows my mind 2021-10-15 03:33:41 it doesn't work if you fully power off and on again, or reboot? 2021-10-15 03:33:52 and what systems have you tested exactly 2021-10-15 03:33:53 Nope! Amazingly. 2021-10-15 03:33:53 riverdc: It's just that I don't want to have the hassle of tagging and I don't do anything too critical to not run edge I think. 2021-10-15 03:34:09 Even resetting UEFI to defaults and reboot doesn't solve it. 2021-10-15 03:36:32 no I think Hello71 is right that it should be avoided, mixing packages like that isn't intended 2021-10-15 03:37:53 if you want to run edge, just run edge. don't make this undebuggable frankensystem 2021-10-15 03:38:39 if you don't know what a soname is, you shouldn't mix alpine releases, and if you do, you still probably shouldn't 2021-10-15 03:39:08 Personally, Edge has been surprisingly stable and easy for me for the past year or so 2021-10-15 03:39:24 Two minor hiccups for a few days each, but nothing untoward. 2021-10-15 03:39:28 I'd go for it 2021-10-15 03:40:07 Nothing worse than what I saw under Sabayon, Solus, or Ubuntu 2021-10-15 03:40:23 And rolling release goodness! 2021-10-15 03:40:56 it depends on what exactly you use. openssl3 has been a moderate disaster 2021-10-15 03:41:02 if you just want rolling, arch is better 2021-10-15 03:41:23 or void 2021-10-15 03:41:27 That's one of the hiccups I mentioned 🤣 2021-10-15 03:41:54 I like the sum of what Alpine brings to the table best so far. 2021-10-15 03:42:05 I distro hop and come right back 2021-10-15 03:42:39 but mixing stable/edge is not recommended, due to ABI differences 2021-10-15 03:42:40 I like it enough I'm trying to learn how to maintain packages I'd like to see and use 2021-10-15 03:43:16 Thanks for the insight everyone, I think I'll be using edge. 2021-10-15 03:44:27 Hello71 i tried Artix since Arch + OpenRC, but it was not the same as Alpine 2021-10-15 03:44:59 There's really good UX here 2021-10-15 03:45:00 well, because only Alpine is Alpine 2021-10-15 03:45:24 Indeed! Void similarly did not impress 2021-10-15 03:45:39 each distro has different goals 2021-10-15 03:45:52 That's fair, yeah 2021-10-15 03:46:25 Lubuntu and then LXLE suited me for a long time 2021-10-15 03:47:08 But I got tired of backlevel everything and Apt being apt 2021-10-15 03:47:59 It's amazing the speed of management under Alpine. It's really effortless and fast 2021-10-15 05:05:55 is it safe to create files in /etc/apk , for example /etc/apk/groups ? 2021-10-15 05:08:33 I would say no, just because apk-tools could decide in the next version to start using that itself 2021-10-15 09:45:26 Any idea why having a proc mount (into a chroot dir) in /etc/fstab doesn't work but running mount -a works? 2021-10-15 14:51:09 anyone here know how to get the consolefont service to set the console font at boot time? I added it to the boot runlevel but strangely it doesn't actually seem to take effect until I log in and restart the service manually 2021-10-15 14:52:18 even shows a "set the font to ..." log message at boot 2021-10-15 14:56:33 hmm, nevermind, setting it to the default runlevel instead of boot seems to accomplish what I want 2021-10-15 15:55:24 hi - can anyone help me with gcc + kernel builds? alpine's bundled gcc doesn't seem to have plugin support, is there a package I'm unaware of or do I have to rebuild gcc and make my own packages? 2021-10-15 16:13:05 you might need the install-base package 2021-10-15 16:13:11 build-base, rather 2021-10-15 16:29:37 yeah there doesn't seem to be plugin support anymore, possibly since the removal of the hardened-kernel package? 2021-10-15 16:37:29 Anyone have a working DE under Edge right now? 2021-10-15 16:38:17 Mine got munged. System is fine otherwise, thankfully. All core services still good, can work-ish from terminal. But I'm lame, so I want gnome-builder so it can hand-hold me in writing rST documentation 2021-10-15 17:32:39 anyone know of cheap laptop thats good as typewriter? 2021-10-15 17:34:01 so, a full size keyboard at least 2021-10-15 17:34:26 with mechanical keys. does a thing exist? 2021-10-15 17:34:38 why mechanical? 2021-10-15 17:34:51 to be "good as typewriter" 2021-10-15 17:35:21 I don't know of any laptop with mechanical keys 2021-10-15 17:35:27 well i just mean that it will be almost only used to type stuff... 2021-10-15 17:35:45 so, a laptop with a keyboard? 2021-10-15 17:36:14 yeah, a good kb 2021-10-15 17:36:33 nvm :p 2021-10-15 17:36:53 so, a good keyboard, like a mec…oh ok 2021-10-15 17:37:28 ikke: there are some gaming laptops that have mech keyboards 2021-10-15 17:37:30 not sure mechanical make a kb better 2021-10-15 17:37:50 i think the msi one iirc 2021-10-15 17:37:51 but some kb u have to hit some keys a speical way for it to register .. 2021-10-15 17:38:05 be nice 2021-10-15 17:38:19 keychron here 2021-10-15 17:38:34 I just have a cheapo HP keyboard here :P 2021-10-15 17:38:46 its new 2021-10-15 17:39:01 must be better than previous one :) 2021-10-15 17:39:16 Heh, well, this is on my desktop :p 2021-10-15 17:39:22 and the new laptop keyboard surely must be better 2021-10-15 17:39:32 ah desktop 2021-10-15 17:41:40 k380 is a nice kb imo 2021-10-15 17:42:53 i also went down the qmk rabit hole tho .... using a nyquist kb too 2021-10-15 17:43:56 but just thinking of a work laptop ... the one i use now gave me a scare by not waking screen, worked after reboot, but made me research a bit just in case 2021-10-15 17:46:07 k380 is nice but not as good as mx keys 2021-10-15 17:46:18 MNT Reform has a mechanical keyboard. 2021-10-15 18:04:48 apk doesn't have any comment syntax for the world file, correct? 2021-10-15 18:05:16 no 2021-10-15 18:05:28 it's not really meant to maintain manually 2021-10-15 18:05:41 apk will probably overwrite it 2021-10-15 18:50:50 If I'd like to tweak the config-edge.x86_64 to include pinctrl_broxton like I did for linux-lts, what would the preffered commit message be? 2021-10-15 18:51:20 I think I understood that any kernel tweaks should have explanation/justification in the MR, but I also think I understood that my commit messages should be pretty minimal 2021-10-15 18:51:49 Saijin_Naib[m]: the subject should be brief, but the body can contain omre 2021-10-15 18:51:59 And what separates them is a newline, right? 2021-10-15 18:52:11 Yes, there is one empty line in between 2021-10-15 18:52:48 Update config-edge.x86_64 - pinctrl_broxton=m 2021-10-15 18:52:48 Add pinctrl_broxton as a module to enable proper platform support for Broxton-based Intel devices (many reference-device based systems with Goldmont/Apollo Lake CPUs or other platforms built around these same CPUs) 2021-10-15 18:52:51 and usually you should make a new line after 50 chars 2021-10-15 18:52:53 Is this acceptable? 2021-10-15 18:53:08 Even in the body? 2021-10-15 18:53:19 No, body is soft wrap around 72 characters 2021-10-15 18:53:41 values vary across internet 😉 2021-10-15 18:54:08 haha, indeed. Every project and sub-project seem to have completely different style guides 2021-10-15 18:54:49 Update config-edge.x86_64 - pinctrl_broxton=m 2021-10-15 18:54:49 Broxton-based Intel devices (many reference-device based systems with Goldmont/Apollo Lake CPUs or other platforms built around these same CPUs) 2021-10-15 18:54:49 Add pinctrl_broxton as a module to enable proper platform support for 2021-10-15 18:55:01 Awh damnit, forgot to keep wrap-checking 2021-10-15 18:55:24 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/COMMITSTYLE.md 2021-10-15 18:59:15 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/EbhMWVlwVytSjBkXkGbHyFuw 2021-10-15 18:59:26 That 50char is tough for subject 2021-10-15 18:59:29 "change config-edge.x86_64 2021-10-15 18:59:41 Saijin_Naib[m]: I use it as a soft limit 2021-10-15 18:59:51 I don't mind going over it, but I still try to keep it brief 2021-10-15 19:00:37 This pass the sniff-test then? 2021-10-15 19:01:02 Saijin_Naib[m]: The package is linux-edge, so that already has most of the context, so I would do: community/linux-edge: enable pinctrl_broxton for x86_64 2021-10-15 19:01:13 or something like that 2021-10-15 19:01:19 Okay, that makes a lot of sense 2021-10-15 19:01:21 Thank you 2021-10-15 19:01:55 community/linux-edge: pinctrl_broxton=m (x86_64) 2021-10-15 19:02:23 Style guide doesn't mention parentheses. Are they good or best avoided? 2021-10-15 19:02:45 We use them where it makes sense 2021-10-15 19:02:53 Mostly for CVEs 2021-10-15 19:03:01 Ah, okay. 2021-10-15 19:03:25 I'll just drop then and put on infront of x86_64 then 2021-10-15 19:03:34 drop them, rather 2021-10-15 19:05:51 Woo, thank you, as always 2021-10-15 19:06:47 Is linux-edge in Community since it isn't "stable" like linux-lts? 2021-10-15 19:07:07 yes 2021-10-15 19:07:38 We do not support it for 2 years 2021-10-15 19:35:28 riverdc: If you want to document why one or more packages are installed, you can use a virtual package for that: apk add -t my-app-dev-deps cmake sdl2-dev sdl2_image-dev 2021-10-15 19:38:54 Oh that's sweet 2021-10-15 19:39:10 So you could for instance mark the baseline install as a virtual package 2021-10-15 19:39:42 You can also break your system with virtual packages, so be careful :) 2021-10-15 19:39:49 (mostly if you mask existing packages) 2021-10-15 19:45:05 Could you like... pin or protect a virtual package from being deleted? 2021-10-15 19:45:30 So back to the fresh-install state, making a virtual package of that and just deleting everything but that to roll back to a known-good working baseline config/install? 2021-10-15 19:57:12 If you try to delete any package that is required by something in world, it will fail. So if you have a virtual package that depends on everything in world after a fresh install, it will do what you're asking for. 2021-10-15 20:03:01 Oh, that's freaking neat 2021-10-15 20:31:43 it won't fail. it will remove it from the world set, which may or may not cause it to be removed from the system 2021-10-15 20:49:02 N8 @ all 2021-10-15 20:51:08 o/ 2021-10-15 21:19:05 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/fqeiwzGjuGKouEtHMluEiBdy 2021-10-15 21:19:12 Should I follow the advice of using --offline here? 2021-10-15 21:19:26 I'm copying the template I used for the lsd rust program APKBUILD 2021-10-15 21:20:16 Question is why it needs to be updated 2021-10-15 21:20:25 Don't know enough about rust to be able to help 2021-10-15 21:22:40 That's a wonderful question. Me either 🤣 2021-10-15 21:22:53 It passes with no --locked or --offline so... Yeah 2021-10-15 21:23:11 yes, but the reason we use --locked is that we want builds to be deterministic 2021-10-15 21:23:13 Trying to get this build/packaged so I can build/package Caesium Image Compressor, which I use daily 2021-10-15 21:23:31 Hello71: You're right, it won't fail, but it will show a warning if it wasn't removed because it's required by something: World updated, but the following packages are not removed due to:" 2021-10-15 21:23:33 Which makes total sense. That may disqualify this package for now since the dev has not tagged any releases yet 2021-10-15 21:23:45 usually 2021-10-16 00:12:52 I'm getting: mount /boot/efi unknown filesystem type 'vfat' on bootup and I'm not real sure what to do. Networking isnt functioning right so I've updated with a chroot but it doesnt seem to make any affect 2021-10-16 00:15:20 the-count: are what stage of booting does the error appear? 2021-10-16 00:15:27 s/are/at/ 2021-10-16 00:17:07 Let me reboot and check 2021-10-16 00:18:52 minimal: do you mind if I share a link to a picture? 2021-10-16 00:21:23 the-count: if it doesn't require javascript... :-) 2021-10-16 00:21:46 Sounds good 2021-10-16 00:25:05 https://kk6mrp.com/content/alpine-boot.jpg 2021-10-16 00:27:19 the-count: right, so its at "Mounting local filesystems", which is when it reads /etc/fstab 2021-10-16 00:27:45 I'm guessing it needs the vfat kernel module and its not loaded 2021-10-16 00:28:55 try adding "vfat" to /etc/modules? 2021-10-16 00:30:52 That doesn't work... also when I run lsmod vfat does not show up either 2021-10-16 00:31:38 you do have vfat module though? find /lib/modules/ -name 'vfat.ko' 2021-10-16 00:33:22 It does show up under the 5.10.72 kernel but the currently booted kernel is 5.10.61 2021-10-16 00:35:26 is this run-from-ram rather than Sys (disk) install? 2021-10-16 00:35:53 no 2021-10-16 00:37:23 at the bottom of your screenshot it says 5.10.61 so if there's no vfat module for that then that's you problem (unless the modules compiled into kernel) 2021-10-16 00:37:58 I just have it installed on my laptop. The long story is that my battery died while updating and then started booting in readonly. In a chroot I was able to finish updating but it might not have updated right since I chrooted from alive iso 2021-10-16 00:38:59 so looks like a newer kernel package installed but somehow the kernel itself didn't get updated 2021-10-16 00:39:24 That seems the most likely, I agree 2021-10-16 00:39:47 its UEFI right, so you're using Grub, in /boot/ you should see the kernel 2021-10-16 00:41:18 Yes, in /boot I see vmlinux-lts... 2021-10-16 00:42:05 if you "grep vmlinux /boot/grub/grub.cfg" you should see the file Grub is using for the kernel, likely to be vmlinuz-virt 2021-10-16 00:43:24 so that much be the old kernel. I guess you may be able to use "apk fix linux-lts" but *note* i've never tried to do that.... 2021-10-16 00:43:32 so can't guarantee what it will do 2021-10-16 00:44:18 There actually isn't any refernces to vmlinux in the grup config 2021-10-16 00:44:42 vmlinuz-lts (that's a "z", not a "x") 2021-10-16 00:45:08 sorry, I typo'd the grep command :-) 2021-10-16 00:46:17 Oh duh... Yes, vmlinuz-lts is in the file 2021-10-16 00:47:35 so the "apk fix" may (or may not) update the kernel file, not sure 2021-10-16 00:47:44 Let me try another chroot and run an apk fix linux-lts 2021-10-16 00:51:31 So it says apk unavailable skipping 2021-10-16 00:53:58 try with "-v" to see more info? 2021-10-16 00:55:40 ... I think that might be because I had the edge repos enabled during thst upgrade 2021-10-16 00:55:56 s/thst/that 2021-10-16 00:56:39 if you're not running Edge its not a good idea to mix Edge with another version in general 2021-10-16 00:57:17 I think that was an uninformed decision at the time 2021-10-16 00:57:31 Is there a way to roll it back? 2021-10-16 00:59:02 Or I guess I could stay on edge and try to fix it as is 2021-10-16 01:00:27 I *think* you can just run "apk fix" (with no package specified) and it will downgrade any necessary packages back to 3.14 (or whatever you were on) but again never had to try that 2021-10-16 01:01:34 apk fix linux-lts works with edge repos re-enabled 2021-10-16 01:03:15 yes but now you have a "hybrid" system - a mix of 3.14 and Edge problems and that can cause problems - e.g. if a program is still the 3.14 version but some of the libraries it depends on have been updated 2021-10-16 01:03:32 s/problems/programs/ 2021-10-16 01:03:55 apk fix doesn't do anything with edge repos disabled 2021-10-16 01:04:59 maybe apk upgrade -a is what I want 2021-10-16 01:05:12 maybe its "apk fix -u" then ("-u" for upgrade which may actually downgrade in your case) 2021-10-16 01:08:36 ok, lets see... 2021-10-16 01:14:39 whoops, tablet went to sleep 2021-10-16 01:28:57 Ok, apk upgrade -a got me fully booting on edge. I need to get going but I'll likely run apk fix -u as you suggested possibly next week 2021-10-16 01:30:52 good luck 2021-10-16 02:00:38 how could we add swi-prolog to the package repository? It is quite a long standing open source project for decades now. 2021-10-16 02:02:12 it builds without any problem: https://github.com/SWI-Prolog/docker-swipl-linux-ci/blob/master/alpine/3.14/Dependencies.docker 2021-10-16 02:09:25 Qa: Do you have any experience packaging software for alpine? 2021-10-16 02:09:48 no, but I have done it for arch in the past 2021-10-16 02:10:20 should not be that different I guess? 2021-10-16 02:11:31 Qa: If you want to try you can follow https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package. The APKBUILD is supposed to be similar to arch's PKGBUILD 2021-10-16 02:12:33 You'll probably want to use https://www.swi-prolog.org/build/unix.html as the instructions and build using the tar ball link on that page 2021-10-16 02:16:40 Do you see any problem with the APKBUILD here? >>> https://github.com/SWI-Prolog/issues/files/6951590/APKBUILD.txt 2021-10-16 02:25:13 Qa: A few small things probably need to be changed. I'll try and get it done soon. (Such as the builddir is hardcoded and the tests have a || true) 2021-10-16 02:26:32 ah, thanks. It is great to have prolog available on alpine. It really surprised me that it was not there already. 2021-10-16 02:27:41 The version there is older, current one is 8.5.0 2021-10-16 02:44:30 Qa: would you be willing to maintain the package? 2021-10-16 02:48:16 sure, but I need to remain anonymous, is that ok? 2021-10-16 02:48:35 use a name generator 2021-10-16 02:48:43 nobody will ask for your birth certificate 2021-10-16 02:49:05 go for a ridiculous but believable one 2021-10-16 03:04:03 Qa: It seems to compile fine, is there any sample code I can use to test it? I have no experience with prolog 2021-10-16 03:05:09 if you run ctest -j8 it does an extensive test of all the features, the one thing may fail is ssl tests because of network problems but that is normal 2021-10-16 03:05:21 if you just want to test 2021-10-16 03:05:30 type write(hello). 2021-10-16 03:05:35 (include the dot at the end) 2021-10-16 03:08:44 you run swipl (the binary) and at the prompt type write(hello). 2021-10-16 03:09:02 Qa: Thanks that seems to work. 2021-10-16 03:09:58 Great. 2021-10-16 03:11:09 Well it turns out it'll be a bit more trouble since the -dev package isn't properly installed so the C headers aren't available. 2021-10-16 03:13:04 I was thinking more about maintaining it, I don't think I can do it 2021-10-16 03:13:26 Strange, it should not have compiled if the headers are not present 2021-10-16 03:13:41 Qa: I meant I'm looking into how to do it. 2021-10-16 03:14:06 ohh maybe you are talking about the ctest -j8; it works fine if you run it from inside the build directory 2021-10-16 03:14:35 What I mean is the SWI-Prolog.h isn't available so swipl-ld doesn't work on https://github.com/PlummersSoftwareLLC/Primes/blob/drag-race/PrimeProlog/solution_1/bitvector.c 2021-10-16 03:15:11 ahh i see. 2021-10-16 03:16:22 normally people don't write C extensions, so they don't need the header file; but I can see what you mean 2021-10-16 03:21:45 If you just want samples you can use this one: https://swish.swi-prolog.org/example/kb.pl and you can find more here: https://swish.swi-prolog.org/example/examples.swinb 2021-10-16 03:24:38 swish is like a jupyter notebook for swi-prolog, you can just paste the examples in file.pl and then run swipl file.pl and type your query at the prompt. In the example you would write e.g. jealous(X,Y). 2021-10-16 03:28:19 Ok I think I've got the -dev package figured out now 2021-10-16 03:30:37 ah great, normally there is not a -dev package for swi-prolog but everything is included in the main one 2021-10-16 03:31:01 like arch linux has swi-prolog and the headers are included in it 2021-10-16 03:31:08 The way alpine does it is that -dev has include files/C libs and -doc has the man pages 2021-10-16 03:31:10 (there is no -dev package) 2021-10-16 03:31:21 ahh I see 2021-10-16 03:31:57 Because you can still run non-C programs without the -dev package 2021-10-16 03:32:07 But the shared libs should be part of the main package because the normal use depends on them 2021-10-16 03:32:47 That's shared libs for other programs to link against, not core functionality libs 2021-10-16 03:34:16 no, the shared libs are loaded at runtime by prolog programs and they are almost always used 2021-10-16 03:34:57 swi-prolog has the ability to load shared libraries at runtime within prolog code and that is how most all the main functionality is achieved 2021-10-16 03:35:10 so if you take the shared libs out the prolog will not function 2021-10-16 03:39:05 I am talking about files like: double_metaphone.so, mallocinfo.so, files.so, archive4pl.so, readutil.so, etc. 2021-10-16 03:39:38 those files are loaded by user prolog code at runtime (without compiling anything) 2021-10-16 03:41:00 It is a very dynamic language, there is not a big division between compilation and running. 2021-10-16 03:41:51 In fact a prolog program is also prolog data (homoiconic); so you can manipulate the prolog program itself as if it was data. 2021-10-16 03:48:37 for example, if you run `read_file_to_string('/tmp/my.txt',Str,[]).` behind the scenes it will load readutils.so at runtime (if it is not loaded). 2021-10-16 03:48:55 This is why I there is no distribution that does a -dev package. 2021-10-16 03:49:41 (it would only have one header file and maybe swipl-ld, because the shared libs need to be included in the main package) 2021-10-16 03:50:50 I'm looking since it seems without -dev files.so is there but readutils.so isn't. I'm not sure if it's a problem with the packaging process 2021-10-16 03:52:06 Oops it's my typo. readutil.so is there 2021-10-16 03:52:34 there should be a *lot* of other shared libs there in the same directory 2021-10-16 03:52:57 I see 51 libs there 2021-10-16 03:53:44 that sounds right 2021-10-16 03:54:37 all those are loaded at runtime, with the exception of libswipl.so which is linked by the swipl executable. 2021-10-16 03:55:00 (they are loaded on an as needed basis, when the code is called). 2021-10-16 03:58:01 what's tipc.so (It's in the arch package but not in my filesystem) 2021-10-16 04:02:23 Qa: I'm not sure if the swipl devs made a mistake but the releases go from 8.3.29 to 8.5.0 2021-10-16 04:02:47 And the arch package is 8.4.0 and the libswipl is also 8.4.0 2021-10-16 04:04:03 no, it's the way the numbering scheme is used 2021-10-16 04:05:01 8.5.0 is the most recent relase with all the features, every once in a while, like every two years, a release with an even number is made but everyone uses the most recent ones 2021-10-16 04:05:31 so you see 8.5.0 just released a little bit ago 2021-10-16 04:06:27 but also 8.4.0 is available which is a release that is updated only like every couple of years, but most everyone uses the odd numbered releases 2021-10-16 04:06:59 so in a couple of years there should be 8.6.0 and 8.7.0 2021-10-16 04:07:11 Oh that makes sense 2021-10-16 04:07:41 but most everyone uses the latest one. 2021-10-16 04:08:31 Qa: So any idea why I don't have mallocinfo.so and tipc.so? Could it be which libraries were installed at build time? 2021-10-16 04:09:49 those are checked at compile time so you are probably missing some dependencies and they were not built 2021-10-16 04:10:06 Any idea which dependencies I need? 2021-10-16 04:10:13 let me see.. 2021-10-16 04:16:40 mallinfo comes from gnu glibc (it's okay if you don't have it, the system adapts) 2021-10-16 04:17:05 I think you use musl that's why you don't have it 2021-10-16 04:17:34 Ok that should be fine then. 2021-10-16 04:18:22 Do you know which deps would be needed for the TIPC module to be built? 2021-10-16 04:18:35 let me look at tipc... 2021-10-16 04:18:52 BTW where are you finding the dep list? 2021-10-16 04:19:26 I am looking at the CMakeLists.txt file and see how the check 2021-10-16 04:20:04 ok tipc checks on linux/tipc.h ; it is a special protocol for highly reliable communications that is included in the linux kernel 2021-10-16 04:20:12 but maybe you configured it out 2021-10-16 04:20:21 that's ok, the system adapts 2021-10-16 04:21:12 Yes it seems like the alpine-virt kernel doesn't have TIPC enabled 2021-10-16 04:21:28 that's okay 2021-10-16 04:25:00 swi-prolog include a distributed key store called paxos which can use tipc if present, but can use other protocols too 2021-10-16 04:25:24 but also includes support for redis and many people use other key stores if needed. 2021-10-16 04:26:06 It is rare because prolog by itself has a 'fact database' so people use key stores in a much less frequent basis than in other languages 2021-10-16 04:40:40 Qa: Ok I see 2021-10-16 04:49:43 have to go, thank you so much for working on it 2021-10-16 04:52:36 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/KJkLWnleoHNReYyytzzwBhDr 2021-10-16 04:52:38 This one is new to me 2021-10-16 04:53:13 Is this the type of thing that requires patching, or did I not setup my APKBUILD properly? 2021-10-16 04:53:13 Official build instructions use plain make with no configure step... 2021-10-16 04:58:32 https://github.com/dfandrich/gpscorrelate/blob/master/INSTALL 2021-10-16 04:59:36 its prefix is hardcoded to /usr/local 2021-10-16 04:59:46 Makefile line 30 2021-10-16 05:00:02 Alright 2021-10-16 05:00:09 So... I need to make a patchfile, right? 2021-10-16 05:00:13 I don't know how to do that 2021-10-16 05:01:23 either that or use prepare() to substitute that one line in the Makefile https://wiki.alpinelinux.org/wiki/APKBUILD_Reference#prepare.28.29 2021-10-16 05:01:42 I'm not sure if that's discouraged in official APKBUILDS or not 2021-10-16 05:03:49 Yeah, I'm looking at this doc 2021-10-16 05:03:54 It seems to indicate my patch goes here, right? 2021-10-16 05:04:37 if you make a patch, then you don't implement prepare(). INstead of you just include a .patch file in the sources 2021-10-16 05:04:43 https://wiki.alpinelinux.org/wiki/Creating_an_Alpine_package#Patches 2021-10-16 05:05:19 Oh, so I can somehow change the Makefile with some code in the prepare block? 2021-10-16 05:05:26 yes 2021-10-16 05:06:56 would export prefix=/usr do it 2021-10-16 05:06:59 Or would the makefile over-ride that? 2021-10-16 05:07:18 the makefile would override that 2021-10-16 05:07:52 mmm 2021-10-16 05:09:29 sed? 2021-10-16 05:09:51 that could work 2021-10-16 05:14:32 shudder. Regex-adjacent 2021-10-16 05:21:19 something like:... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/RsKTGWtplRtcuuMWYssfTahx) 2021-10-16 05:23:16 got the argument order wrong -_- 2021-10-16 05:24:55 raven523: Thank you so much! 2021-10-16 05:25:53 CLI works 2021-10-16 05:26:04 GUI throws an error, but I can't tell if it is "real" or not 2021-10-16 05:26:31 Saijin_Naib[m]: You might want to use sed with pipes as the separator: sed -i Makefile 's|= /usr/local|= /usr|' 2021-10-16 05:26:42 Is that safer? 2021-10-16 05:26:45 Or whatever 2021-10-16 05:26:48 That way you don't need backslashes 2021-10-16 05:26:54 So it's more readable 2021-10-16 05:27:16 none of this is readable to me 🤣 2021-10-16 05:27:53 Thank you, adjusted 2021-10-16 05:35:32 If this package has both a CLI and a GTK+3.0 GUI, do I need to put GTK+3.0 as a depends? 2021-10-16 05:35:44 Since conceivably, someone could use it CLI-only and be 100% good 2021-10-16 05:37:32 Error:... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/QcWCQhhrmkBNCaCzCbaGrpcQ) 2021-10-16 08:16:44 hello! i need to run a script on system shutdown, but not on reboot. i don't care all that much how this is called (could be via a /etc/local.d/something.stop script or an init script). i've checked environment variables to see if there is anything that let's me distinguish the two, but couldn't find anything. any ideas how i can distinguish a shutdown from a reboot? 2021-10-16 08:18:21 wej: the absolutely filthy and disgusting approach would be to use a custom shutdown command, if nothing else materializes 2021-10-16 08:18:57 rcf: yeah, i thought about that also, but really didn't like the idea ^^ 2021-10-16 08:19:29 I usually reserve that for embedded systems where I can hide away the shame 2021-10-16 08:20:09 :D 2021-10-16 08:20:37 Of course I suppose one should also ask: what are you trying to do that is so undesirable to do on a reboot? 2021-10-16 08:21:21 rcf: i have a raspberry pi that i want to power off externally on shutdown 2021-10-16 08:21:38 it wouldn't be all that helpful if it would happen on reboot also 2021-10-16 08:22:02 How are you triggering the external poweroff? 2021-10-16 08:22:27 via a command over serial to an attached micro controller 2021-10-16 08:23:14 Do your reboots usually take a consistent amount of time? 2021-10-16 08:24:13 i guess so, but i wouldn't bet on it being always exactly the same 2021-10-16 08:27:48 Depending on your use case, you could set the hard poweroff to operate on a timer, which is cancelled by another command sent during boot. 2021-10-16 08:28:35 it works on a timer already, so it doesn't immediately turns the power off, so that would be possible 2021-10-16 08:28:43 but it feels kind of fragile 2021-10-16 08:46:13 As long all filesystem are unmounted or remounted read-only it should be OK to just cut power. 2021-10-16 08:52:07 ccx^xmpp: yes, that's true, but i was thinking if it takes longer than expected to reboot, i wouldn't be able to cancel the external poweroff 2021-10-16 08:56:16 Sure. I'm not familiar with busybox init but with sysv you should be able to either define stuff for reboot runlevel or replace the poweroff command which is called to power off. 2021-10-16 08:57:20 It's probably also feasible with s6-linux-init but haven't looked too deeply into it's runlevels yet. 2021-10-16 09:00:28 it looks like there is only one runlevel "shutdown" which is used to both reboot and shutdown, but replacing the poweroff command could work 2021-10-16 09:00:54 i was hoping i wouldn't have to replace files that might be overwritten by a system update 2021-10-16 09:02:18 Yeah, that is unpleasant. 2021-10-16 09:03:08 But I guess that's just the limitation of busybox init. 2021-10-16 09:09:27 yeah 2021-10-16 09:09:35 anyway, thanks for your help :) 2021-10-16 09:15:20 Saijin_Naib[m]: To your last question: that's what subpackages are for 2021-10-16 09:16:08 You can split into multiple packages, and only the package with the gui components would depend on gtk 2021-10-16 10:39:09 since its world browser day ... what browser do you use? is there a text browser that can handle js/netbank? 2021-10-16 10:44:01 probably not 2021-10-16 10:44:08 js is hard to do anyway 2021-10-16 10:44:15 doing it on a terminal? 2021-10-16 10:44:28 i thibk stuff like edbrowse has js support 2021-10-16 10:44:51 netsurf isnt textual but it can do miminal js and can do framebuffer 2021-10-16 10:44:58 edbrowse is well, ed 2021-10-16 10:47:05 netsurf is not maintained i think 2021-10-16 10:47:30 didnt know about edbrowse 2021-10-16 11:26:41 netsurf is maintained iirc 2021-10-16 11:26:46 it got a plan 9 pot recently 2021-10-16 11:26:48 port 2021-10-16 11:26:49 so 2021-10-16 11:27:01 at least something must be goin on 2021-10-16 11:46:17 6 2021-10-16 12:17:53 Yeah, I remember their IRC being reasonably active. 2021-10-16 12:19:42 The main reason I don't use netsurf actively is doesn't suppoer keyboard-only usage. 2021-10-16 12:32:28 ccx^xmpp: what do you use? 2021-10-16 12:42:23 Firefox, vimb, qutebrowser, w3m 2021-10-16 12:45:52 vimb just webcrashes for me... 2021-10-16 12:58:07 ACTION should change rpi4 case, small 3 heatsinks in closed steel case arent enough when compiling kernel :/ 82 C so throttling hard 2021-10-16 13:17:49 raven523, Saijin_Naib[m]: you are using make wrong. you need to run "make prefix=/usr", not "prefix=/usr make" 2021-10-16 14:02:34 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/LIqsnMtvOcXqCPufZKQzDWUP 2021-10-16 14:03:56 in the makefile it looks like this: 2021-10-16 14:03:56 https://github.com/dfandrich/gpscorrelate/blob/master/Makefile 2021-10-16 14:04:58 Does abuild over-ride the CFLAGS/LDFLAGS this Makefile sets, or do I need to sed those out? 2021-10-16 14:08:54 A good Makefile will use CLFLAGS/LDFLAGS from the environment 2021-10-16 14:11:35 This one might not be very good, haha 2021-10-16 14:11:40 It seems everything is hard-coded 2021-10-16 14:13:37 problem is posix make doesn't have ?= 2021-10-16 14:14:42 but it is easy to do make CC=$CC CFLAGS=$CFLAGS 2021-10-16 14:17:55 Hmm, okay. Do i need to adjust the other lines at the top like CXX, EXEEXT? 2021-10-16 14:38:55 Oh wow I broke it bad haha 2021-10-16 18:02:52 oh hi! 2021-10-16 18:06:27 how to find out who is compiling the package ntpsec for alpine? 2021-10-16 18:07:03 I got some security related issue with how the ntpsec package is compiled, already put up on gitlab but after a week or so still no response: https://gitlab.alpinelinux.org/alpine/aports/-/issues/13076 2021-10-16 18:08:10 https://pkgs.alpinelinux.org/package/edge/community/x86_64/ntpsec lists it has no maintainer 2021-10-16 18:11:31 It's just a matter of adding libcap-dev to makedepends 2021-10-16 18:12:03 I know how to manually fix it but would be nice if the official repo got the update too 2021-10-16 18:12:11 yes 2021-10-16 18:12:14 "libcap (under Linux) enables dropping root and is not strictly required." 2021-10-16 18:12:22 also changing so that default startup includes -U 123:123 2021-10-16 18:12:40 but its required to droproot 2021-10-16 18:12:43 yup 2021-10-16 18:12:50 and having ntp being runned as root is just as bad as it can get :) 2021-10-16 18:13:17 And I suppose it uses capabilities then to be able to change the system tiome 2021-10-16 18:13:36 probably but didnt see it complain about that 2021-10-16 18:13:49 it starts and works happily as root 2021-10-16 18:14:08 but when trying to run it as ntp/ntp (-U 123:123) is just crashes upon start 2021-10-16 18:14:09 as root, it can change the time, but if it drops root, it can no longer do it by default 2021-10-16 18:14:15 since libdep is missing in its linking 2021-10-16 18:14:20 libdep? 2021-10-16 18:14:28 or libcap?> 2021-10-16 18:14:30 err libcap 2021-10-16 18:15:26 There is no service included 2021-10-16 18:15:59 ? 2021-10-16 18:16:22 The package does not include an openrc init file to have it run as a service 2021-10-16 18:17:43 I for sure got an ntpd in /etc/init.d 2021-10-16 18:17:51 looks like an openrc-run to me 2021-10-16 18:18:11 Are you sure that's ntpsec? 2021-10-16 18:18:30 there is no other ntp in the packages 2021-10-16 18:18:35 or ehm 2021-10-16 18:18:41 https://pkgs.alpinelinux.org/contents?file=&path=&name=ntpsec*&branch=edge&arch=x86_64 2021-10-16 18:18:43 could it be some leftover from the openntpd ? 2021-10-16 18:18:46 or ntp-client? 2021-10-16 18:18:55 Yes. That package overwrites /usr/sbin/ntpd from whatever other ntpd 2021-10-16 18:19:16 Apachez: busybox provides ntpd 2021-10-16 18:19:29 mercenary: right 2021-10-16 18:19:54 I know I originally installed openntpd and then removed it 2021-10-16 18:19:57 and then installed ntpsec 2021-10-16 18:20:21 so I guess that needs to be fixed aswell then? 2021-10-16 18:20:29 So ntpsec reuses the default ntpd service as it masks bb ntpd 2021-10-16 18:20:40 https://pkgs.alpinelinux.org/contents?file=ntpd&path=%2Fetc%2Finit.d&name=&branch=v3.14&arch=x86_64 2021-10-16 18:21:55 Apachez: the downside of that is that we cannot add options that only work for ntpsec 2021-10-16 18:23:28 for command_args the content of /etc/init.d/ntpd should read command_args="${NTPD_OPTS}" and NTPD_OPTS is set in /etc/conf.d/ntpd and should read something like: NTPD_OPTS="-g -N -c /etc/ntp.conf -u 123:123" 2021-10-16 18:24:06 but that can be manually changed but would be sane to default user ntp and group ntp rather than user root 2021-10-16 18:24:33 as long as linking libcap is included in the package build :) 2021-10-16 18:24:40 Apachez: ntpd: unrecognized option: u 2021-10-16 18:25:17 [-u user[:'group']] 2021-10-16 18:25:26 do you have ntpsec ? 2021-10-16 18:25:51 https://tpaste.us/w1Mv 2021-10-16 18:26:17 yes thats the busybox edition 2021-10-16 18:26:20 correct 2021-10-16 18:26:23 that seems to be lacking droproot 2021-10-16 18:26:30 but these options _have_ to work with bb ntpd 2021-10-16 18:26:45 Im talking about the ntpsec one who should have droproot but is missing linking to libcap so I cannot start it with -u 123:123 which ntpsec supports 2021-10-16 18:26:51 yes, I understand 2021-10-16 18:27:03 again, Im happy to manually alter the /etc/conf.d/ntpd 2021-10-16 18:27:12 but I cant right now because if I do then ntpsec wont start 2021-10-16 18:27:15 but what you are maybe missing is that if you don't install ntpsec, you have bb ntpd, and the ntpd service will try to start bb ntpd 2021-10-16 18:27:35 Apachez: I will shortly push an update that includes libcap 2021-10-16 18:28:04 But we cannot adjust the confd file to use options that only work for ntpsec 2021-10-16 18:28:20 So yes, if you want that, you have to do that yourself on your system 2021-10-16 18:28:36 as long as libcap is included one can alter this manually... but it would still be nice if ntpsec had its own /etc/conf.d/ntpd where -u 123:123 is included... so either you run the busybox ntpd as is or you install the "real" ntpsec who then will overwrite some common files 2021-10-16 18:28:47 its like with other busybox commands 2021-10-16 18:29:07 Yes, but for commands there is a provision, but not for service files 2021-10-16 18:29:22 either you accept the symlink towards busybox builtin stuff and if you dont like it you install the "real" package who put a binary instead of that symlink and "overruling" whatever busybox comes with 2021-10-16 18:30:31 once a ntpsec that includes libcap is available through official repos... can I install just that into my 3.14.2 install or do I have to change repos into edge ? 2021-10-16 18:30:45 because I assume it wont be backported into 3.14.2 ? 2021-10-16 18:31:17 I suppose this can be backported to 3.14 2021-10-16 18:31:30 but would be nice if you could at least test it on edge first 2021-10-16 18:32:06 let me know when its available and I can switch to edge repo for this (virtual) install 2021-10-16 18:32:48 enabled edge main, community and testing 2021-10-16 18:35:24 Apachez: pushed, it should be available shortly 2021-10-16 18:35:47 "Welcome to Alpine Linux 3.15.0_alpha20210804 (edge)" :) 2021-10-16 18:43:32 Im guessing a regular apk update && apk upgrade would fetch that? 2021-10-16 18:45:04 yes 2021-10-16 18:50:54 I guess there will be some delays for the mirrors aswell? is there a central repo I could point to in this case? 2021-10-16 18:52:14 https://dl-cdn.alpinelinux.org/alpine/edge/community/x86_64/ already has it 2021-10-16 19:02:40 tnx 2021-10-16 19:02:41 (1/2) Upgrading ntpsec (1.2.1-r1 -> 1.2.1-r2) 2021-10-16 19:02:41 (2/2) Upgrading ntpsec-doc (1.2.1-r1 -> 1.2.1-r2) 2021-10-16 19:04:38 a regular start without -u looks good: Oct 16 21:03:45 alpine daemon.info ntpd[2678]: DNS: dns_take_status: ntp.se=>good, 0 2021-10-16 19:05:11 ntpq -p 127.0.0.1 validates that it got stratum 1 from ntp.se (which I configured to be used) 2021-10-16 19:05:18 now lets see how the -u part works 2021-10-16 19:06:03 it started 2021-10-16 19:06:05 2875 ntp 0:00 /usr/sbin/ntpd -4 -g -N -c /etc/ntp.conf -u 123:123 2021-10-16 19:06:41 output from messages looks good too (complains about that /var/NTP doesnt exist so it cannot store stats but I can fix that manually) 2021-10-16 19:06:58 ntpq -p 127.0.0.1 validates that it got stratum 1 from ntp.se 2021-10-16 19:10:15 and synking from another host towards the host with alpine running 1.2.1-r2 worked too 2021-10-16 19:14:58 from my point of view it seems to be working as expecting 2021-10-16 19:15:18 should other fixes be added aswell like the missing /var/NTP who should have owner ntp:ntp ? 2021-10-16 19:17:31 would be great if at least the libcap fix could be backported into 3.14.2 :) 2021-10-16 19:20:03 to sum it up regarding ntpsec 1.2.1-r2: it now start with user ntp and group ntp, I have verified that it successfully sync time with external source (verified against ntp.se) and I have also verified that a client can sync its time towards the alpine host with ntpsec 1.2.1-r2. Verified by using "ntpq -p 127.0.0.1" so check the local ntpd status (both on alpine and on the client syncing against alpine) 2021-10-16 19:25:13 Apachez: if you feel up to it, you could make an MR 2021-10-16 19:25:19 even claim maintainershup 2021-10-16 19:25:23 maintainership* 2021-10-16 19:28:35 I lack too much knowledge of that unfortunately 2021-10-16 19:30:13 so I would be glad if somebody else could fix -r2 into 3.14.2 :) 2021-10-16 19:30:15 you had the knowledge of how to fix the capability and (probably) know how to edit the apkbuild to even make the change 2021-10-16 19:30:22 that is good enough 2021-10-16 19:30:49 but the apkbuild is already changed in edge? 2021-10-16 19:30:55 just sign up to upstream notifications for updates and that's all you will need to worry about :) 2021-10-16 19:30:57 yes, it is 2021-10-16 19:30:57 cant that just be copied into main of 3.14.2 ? 2021-10-16 19:31:05 i mean if you had to make it, i'm not commenting on the backport 2021-10-16 19:31:13 i'm commenting on maintaining it 2021-10-16 19:31:16 ikke made the magic 2021-10-16 19:31:19 i know 2021-10-16 19:31:29 i'm saying you can totally maintain it :) 2021-10-16 19:31:45 but then it will be my fault when stuff breaks ;) 2021-10-16 19:31:58 https://git.alpinelinux.org/aports/commit/?id=b56c0351e62c 2021-10-16 19:32:01 this was the change 2021-10-16 19:32:01 Im more comfortable to just file bug reports and feature requests ;) 2021-10-16 19:32:08 harder to break than you think ;) 2021-10-16 19:32:39 Apachez: and yes, backporting it to 3.14 is a matter of cherry-picking the commit 2021-10-16 19:34:09 well I guess its time to learn then ;) 2021-10-16 19:34:33 cant that be done centralized? 2021-10-16 19:34:53 or do I have to commit the same commit myself into the 3.14.2 branch, I have zero clue on how that works 2021-10-16 19:35:05 again Im just into bug reporting and feature requesting ;) 2021-10-16 19:35:22 Apachez: each release is a branch in git 2021-10-16 19:35:37 edge tracks master, v3.14 tracks 3.14-stable 2021-10-16 19:37:21 but I guess only admins should be able to "lets take the commit of b56c0351e62c and apply that into the same file of 3.14 and have it built" ? 2021-10-16 19:37:39 as in it shows up in the official repo of 3.14 2021-10-16 19:37:43 Apachez: there are developers who have push access 2021-10-16 19:39:34 you can make the mr for it yourself pretty sure, someone always has to review it if you don't have access 2021-10-16 19:40:18 to backport it you checkout 3.14-stable and git cherry-pick the commit in edge onto it, then mr that 2021-10-16 19:41:06 never done that before... would need to hold somebodys hand for that =9 2021-10-16 19:41:07 =) 2021-10-16 19:43:28 holding hands is very cute 2021-10-16 19:46:13 Forking in progress... 2021-10-16 19:51:14 ok so there is now a MR https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/26495 2021-10-16 19:56:31 Made some remarks 2021-10-16 19:56:34 hmm... do I have to change something in my MR to reflect that commit msg? 2021-10-16 19:57:12 You can amend the original commit and squash that 'suggestion' commit into it 2021-10-16 19:59:34 so something like this then? 2021-10-16 20:00:09 You need to change the actual commit, which you can do locall 2021-10-16 20:00:10 locally 2021-10-16 20:00:17 first fetch your branch (to get the suggestion commit) 2021-10-16 20:00:50 I did it originally through the browser 2021-10-16 20:00:57 so just changing the description isnt enough? 2021-10-16 20:01:17 AH ok, no. That's only kept in the merge request 2021-10-16 20:01:49 If you look here: https://gitlab.alpinelinux.org/alpine/aports/-/commits/master, so you see each commit adheres to that commit style 2021-10-16 20:02:43 yeah but I edited the title and description for this merge request, isnt that enough? 2021-10-16 20:02:51 no 2021-10-16 20:02:55 it's the commit message itself 2021-10-16 20:03:23 We have a flat commit style, we do not merge the merge request (where a merge commit is created) 2021-10-16 20:03:45 but we rebase each commit and then 'fast-forward' merge master 2021-10-16 20:04:41 at first I was "oh that was easy" and now its like "Im too slow this evening" :) 2021-10-16 20:05:03 Through the webinterface you can only do so much 2021-10-16 20:05:04 so my merge request is like accepted and done? 2021-10-16 20:05:07 If you want, I can fix it up for you 2021-10-16 20:05:21 but I still need to create a new commit? 2021-10-16 20:05:31 na its fine, time to learn... 2021-10-16 20:05:52 ok 2021-10-16 20:05:53 even if I currently have zero clue of what Im supposed to do 2021-10-16 20:06:18 So, because the webinterface does not allow you to edit commits, you have to clone your fork locallyu 2021-10-16 20:08:39 there is an open in web ide? 2021-10-16 20:09:01 Does not suffice 2021-10-16 20:09:05 you can only make new comits 2021-10-16 20:09:08 commits 2021-10-16 20:09:16 while the goal is to 'edit' the existing commits 2021-10-16 20:10:15 and cloning the fork is done locally in cli I guess? 2021-10-16 20:10:22 yes 2021-10-16 20:12:19 Hello everyone. I have a problem mounting devices. 2021-10-16 20:12:53 I tried googling but nope,, I dont get how to do that 2021-10-16 20:13:29 I am using alpine linux without any de or wm. I didnt found any solution on google / stackexchange or the alpine wiki 2021-10-16 20:13:38 Apachez: git clone git@gitlab.alpinelinux.org:Apachez-/aports.git 2021-10-16 20:14:05 what problem for mounting devices do you have 2021-10-16 20:15:12 when i enter "doas mount /dev/sdc/ /mnt" or "doas mount /dev/sdc/ /foo/bar/" it says "mount: mounting /dev/sdc/ on /mnt failed: Not a directory" 2021-10-16 20:16:25 The same happens with any directory. fdisk tells me that sdc is the right name for the usb device. 2021-10-16 20:17:07 i tested different usb sticks with different filesystems but this does not help either 2021-10-16 20:17:26 ikke: I suppose I should change that into @gitlab... ? 2021-10-16 20:17:37 Apachez: no, for ssh, the username is always git 2021-10-16 20:17:46 but maybe https is easier for you 2021-10-16 20:18:13 git clone https://gitlab.alpinelinux.org/Apachez-/aports.git 2021-10-16 20:18:35 I also tried "mount /dev/sdc/ /mnt" as root but same problem. 2021-10-16 20:18:46 humanbeing: does /mnt exist? 2021-10-16 20:19:39 yes /mnt exists, i also created some directory ~\test with mkdir and got the same error when trying to mount in ~\test 2021-10-16 20:20:29 * ~/test 2021-10-16 20:20:58 humanbeing: anything in dmesg? 2021-10-16 20:21:20 what are the contents of /etc/fstab? 2021-10-16 20:22:10 ok cloned, now what? :) 2021-10-16 20:22:43 git checkout Apachez--3.14-stable-patch-38884 2021-10-16 20:23:30 fatal: not a git repository (or any of the parent directories): .git 2021-10-16 20:23:39 oh, first cd into the aports dir 2021-10-16 20:23:47 I need to traverse into the dir with APKBUILD? 2021-10-16 20:23:57 no, just the git repo dir is enough 2021-10-16 20:24:01 ok 2021-10-16 20:24:02 done 2021-10-16 20:24:10 you checked out that branch? 2021-10-16 20:24:19 yeah? 2021-10-16 20:24:24 ok 2021-10-16 20:24:25 Branch 'Apachez--3.14-stable-patch-38884' set up to track remote branch 'Apachez--3.14-stable-patch-38884' from 'origin'. 2021-10-16 20:24:26 Switched to a new branch 'Apachez--3.14-stable-patch-38884' 2021-10-16 20:24:30 yes, good 2021-10-16 20:24:35 @ikke: /etc/fstab contains three lines starting with 'UUID=' and then followed by "/" "/boot" and "swap" as well as a line with "/dev/cdrom /media/cdrom ...." and a line with "/dev/usbdisk /media/usb ..." 2021-10-16 20:25:00 Apachez: now git rebase -i origin/3.14-stable 2021-10-16 20:25:21 still in aports dir? 2021-10-16 20:25:24 Apachez: yes 2021-10-16 20:25:39 humanbeing: anything referring to the devices you want to mount? 2021-10-16 20:25:39 ikke: I am not familiar with dmesg but in the last lines it outputs it recocnizes my usb 2021-10-16 20:27:30 I got a git-rebase-todo 2021-10-16 20:27:53 good 2021-10-16 20:28:01 two lines of pick 2021-10-16 20:28:04 eys 2021-10-16 20:28:04 and then some green text 2021-10-16 20:28:17 so it's about those first 2 lines 2021-10-16 20:28:24 the 2nd line should be the suggestion commit, correct? 2021-10-16 20:28:30 yup 2021-10-16 20:28:36 change pick to fixup there 2021-10-16 20:28:42 ikke: the content of /etc/fstab does not change after unplugging the usb so i guess no 2021-10-16 20:28:51 humanbeing: it would not change 2021-10-16 20:28:52 so first line leave as pick, and change 2nd line from pick into fixup? 2021-10-16 20:29:00 ikke: but fdisk -l lists my usb 2021-10-16 20:29:01 the first line should be reword 2021-10-16 20:29:12 so pick -> reword 2021-10-16 20:29:29 yeah now I got 1st as reword and 2nd as fixup 2021-10-16 20:29:32 good 2021-10-16 20:29:35 save and quit 2021-10-16 20:30:14 like exit? 2021-10-16 20:30:18 yes 2021-10-16 20:30:36 file name to write: sorry 2021-10-16 20:30:49 just enter there 2021-10-16 20:30:51 it was a scroll 2021-10-16 20:32:58 once you save that file (should be 'todo'), it should open up an editor again with the commit message of the first commit 2021-10-16 20:33:07 can you even mount a /dev/sdc? i'm pretty sure it normally has /dev/sdc1 /dev/sdc2 etc 2021-10-16 20:33:28 @ikke: the error message that mount outputs also recocnizes the existence of /dev/sdc/ but claims that /mnt or /foo/bar is not a directory. 2021-10-16 20:33:32 psykose: /dev/sdc could be a proper partition 2021-10-16 20:33:39 i think it's possible if it's a non-partitioned partitioned or whatever it is, but i think the device is just not read properly 2021-10-16 20:33:40 yeah 2021-10-16 20:33:59 Especially if you want the partition to be recognized by windows 2021-10-16 20:34:20 psykose: the same thing happens with dev/sdc1 and /dev/sdc2 2021-10-16 20:34:27 humanbeing: the error is apparently a different error then when the directory you want to mount onto does not exist 2021-10-16 20:34:35 what filesystem do you have on the device 2021-10-16 20:35:00 afaik i get the same error if i don't modprobe the fs module before mounting something that needs it 2021-10-16 20:35:21 bb mount does not automatically modprobe iirc 2021-10-16 20:35:25 yep 2021-10-16 20:35:46 a pastebin of fdisk -l would be nice too 2021-10-16 20:36:01 ikke: fat32 as gparted on my ubuntu machine tells me 2021-10-16 20:36:39 I have amended to fix the commit msg 2021-10-16 20:36:57 Successfully rebased and updated refs/heads/Apachez--3.14-stable-patch-38884. 2021-10-16 20:37:02 am I done now? :) 2021-10-16 20:37:05 you could try modprobe fat 2021-10-16 20:37:20 and post the fdisk -l output just to make sure on something 2021-10-16 20:37:24 Apachez: so 'git log origin/3.14-stable..' returns just one commit with the expected message? 2021-10-16 20:38:25 there are a few commits there 2021-10-16 20:38:34 did you include the dots? 2021-10-16 20:39:56 now I did :) 2021-10-16 20:39:58 commit 7c85ad665354f6c3dd1efde3f0846c1070a2e67e (HEAD -> Apachez--3.14-stable-patch-38884) 2021-10-16 20:40:11 ok 2021-10-16 20:40:12 looks ok, same msg as the updated merge request 2021-10-16 20:40:22 now you can push this commit: git push --force-with-lease origin 2021-10-16 20:41:50 done 2021-10-16 20:43:07 Good: https://gitlab.alpinelinux.org/Apachez-/aports/-/compare/3.14-stable...Apachez--3.14-stable-patch-38884?from_project_id=1505 🙂 2021-10-16 20:44:04 and now Im done? :) 2021-10-16 20:44:17 I guess I can remove the local git cache without breaking stuff ? 2021-10-16 20:44:36 yes, you can delete the repo 2021-10-16 20:44:44 and congratulations to your first git commit in aports :)( 2021-10-16 20:44:56 tnx for taking your time to guide :) 2021-10-16 20:45:03 No problem 2021-10-16 20:45:15 psykose: not sure what exactly do you mean by pastebin, but fdisk -l outputs "Device: /dev/sdc1" "Boot: " "StartCHS: 0,37,14" "StartLBA: 2344" "EndLBA: 15585279" "Sectors: 15582936" "Size: 7608M" "Id: b" "Type: Win95 FAT32" 2021-10-16 20:46:00 what happens next? will it take a day or two before enough of developers/admins approved that commit or will it go faster than that to show up in the repository as a new apk ? 2021-10-16 20:46:33 This will now be built by the builders and in about 15-30 minutes it should be available :) 2021-10-16 20:47:09 https://tpaste.us/Jqa0 2021-10-16 20:47:23 bad gateway 2021-10-16 20:47:34 F5 2021-10-16 20:47:55 "doas modprobe fat" outputs nothing 2021-10-16 20:49:43 and now if you mount /dev/sdc1 /mnt what happens 2021-10-16 20:50:25 obviously you cannot mount device /dev/sdc/ because /dev/sdc is a file, not a directory 2021-10-16 20:51:03 "doas mount /dev/sdc1 /mnt" outputs "mount: mounting /dev/sdc1 on /mnt failed: Invalid argument" 2021-10-16 20:51:35 "doas mount /dev/sdc1/ /mnt" outputs "mount: mounting /dev/sdc1/ on /mnt failed: Not a directory" 2021-10-16 20:52:08 Hello71: huh? 2021-10-16 20:53:07 humanbeing: dmesg | nc termbin.com 9999 2021-10-16 20:54:16 I guess https://pkgs.alpinelinux.org/ is somewhat realtime or is there some delay between that and whats compiled? 2021-10-16 20:54:53 There is a job that runs every 15 minutes, to sync the packages from the mirror 2021-10-16 20:56:46 does "makedepends" include the actual library on a fresh install? 2021-10-16 20:56:51 https://termbin.com/g9wi 2021-10-16 20:57:08 Apachez: abuild traces the runtime dependencies and automatically adds them as dependencies 2021-10-16 20:57:12 I mean we added libcap-dev as makedepends... but would that include "libcap" once the ntpsec package is installed? 2021-10-16 20:57:18 ahh ok 2021-10-16 20:58:10 https://tpaste.us/8E7K 2021-10-16 20:58:26 so:libcap.so.2 2021-10-16 20:58:51 and for fdisk -l: termbin.com/fj1b 2021-10-16 21:02:52 Hello71: does this help? I have no experience with dmesg, so i have no clue what those lines mean 2021-10-16 21:03:32 what happens if you run doas mount -t vfat /dev/sdc1 /mnt 2021-10-16 21:03:39 you got both "dmesg" and "/var/log/messages" to dig through 2021-10-16 21:04:01 I don't see the actual block devices being created 2021-10-16 21:05:06 Hello71: this worked!! :) 2021-10-16 21:07:08 Hello71: thank you and all the others for your time and help 2021-10-16 21:07:46 Hello71: So this is the command i will have to run everytime to mount this usb? 2021-10-16 21:09:27 humanbeing: or you install util-linux-misc 2021-10-16 21:10:40 ikke: how is the apk called? "apk add util-linux-misc" says that there is no such package 2021-10-16 21:10:54 humanbeing: what version of alpine are you on? 2021-10-16 21:11:05 might be just util-linux 2021-10-16 21:11:45 ikke: x86_64 2021-10-16 21:11:58 cat /etc/alpine-release 2021-10-16 21:12:48 ikke: 3.14.2 2021-10-16 21:12:49 that's only on edge no? 2021-10-16 21:12:54 psykose: yes 2021-10-16 21:12:59 humanbeing: so util-linux 2021-10-16 21:14:20 Thanks, i installed util-linux and now mount works fine without the -t flag 2021-10-16 21:15:11 or echo vfat > /etc/filesystems 2021-10-16 21:15:42 Hello71: shall i do this additionally? 2021-10-16 21:16:08 you need to do one or more of apk add util-linux or echo vfat > /etc/filesystems 2021-10-16 21:16:15 ikke: and now its live on the pkgs aswell, thanks for all the help! :) 2021-10-16 21:16:26 Apachez: no problem 2021-10-16 21:18:10 Hello71: /etc/filesystems does not exist, i will create it now and add "vfat" as the only line, or will this cause problems? 2021-10-16 21:18:27 that's fine 2021-10-16 21:18:45 if you want to save some disk space you can do that and then delete util-linux 2021-10-16 21:18:51 otherwise you can just use util-linux 2021-10-16 21:19:24 ikke: can you check up on that mips64 version? was it just the synk towards the webserver of pkgs that is missing -r1 or did the build fail on mips64? 2021-10-16 21:19:29 https://pkgs.alpinelinux.org/packages?name=ntpsec&branch=v3.14 2021-10-16 21:19:33 maybe util-linux is better if you are not aware of the limitations of bb 2021-10-16 21:19:37 Hello71: Thank you very much. Your help was really useful. 2021-10-16 21:19:50 Apachez: The mips builder is awol, and probably will not come back 2021-10-16 21:19:57 ok 2021-10-16 21:20:16 The architecture is basically EOL 2021-10-16 21:20:38 ikke: yes i will keep util-linux installed. I run alpine on an old machine to lear how to do everything in the terminal but i am only a beginner so it linux-utils looks useful 2021-10-16 21:24:11 I will leave now. I learned a lot 2021-10-16 21:24:39 o/ 2021-10-16 21:36:57 Having "Illegal Instruction" probs with onboard, Firefox... (alpine edge) looks like it's posted as an issue. Just feelin' down. 2021-10-16 21:37:33 so for whatever reason I accidently created a file named "-g" in the root of alpine installation... how do I remove that? :) 2021-10-16 21:37:47 rm \-g doesnt work rm -g doesnt work rm "-g" doesnt work 2021-10-16 21:38:05 rm '-g' didnt work either 2021-10-16 21:38:14 rm -- -g 2021-10-16 21:38:22 tnx :) 2021-10-16 21:38:36 after -- no options are parsed anymore 2021-10-16 21:38:50 counts for many linux commands 2021-10-16 21:39:33 rm /-g is better 2021-10-16 21:39:40 or rm ./-g 2021-10-16 21:40:15 why better? 2021-10-16 21:42:36 afaik there are some things that don't implement -- 2021-10-16 21:42:49 but for ./-g to fail they need to have straight up insane argparsing 2021-10-16 21:43:12 not for any common utilities of course 2021-10-16 21:44:51 hm, i thought -- wasn't posix, but it is at least in 2018 2021-10-16 21:45:24 but yes, -- should be supported for rm, but ./ is usable in more programs 2021-10-16 22:15:54 ikke: as mentioned in libera.chat, how come python3 is on the dependency list of ntpsec ? 2021-10-16 22:17:51 Apachez: At least some parts are implemented in python 2021-10-16 22:18:58 i forgot the libera version existed, hah 2021-10-17 03:30:15 Anyone having trouble with gtk apps under Edge? 2021-10-17 03:31:19 I think it was my libffi or icu-libs getting upgraded, I think, but about the only thing I can barely use now is lxqt. Firefox, flatpak, Geary, GNOME Builder, et all crash with illegal instruction 2021-10-17 03:31:36 I imagine they all must be rebuilt against the new libraries to not fail? 2021-10-17 03:32:25 Terminal session is good, though dbus appears to be dead as things like lxpolkit and lxpowermanager complain about not being able to talk to dbus 2021-10-17 03:32:53 icu-libs is used in more things than libffi 2021-10-17 03:38:06 Perhaps that is what it was then 2021-10-17 03:38:12 I guess I'm rather stuck, right? 2021-10-17 03:38:28 Mostly don't care, but I really lean on GNOME Builder, like, a lot 2021-10-17 03:38:32 That being gone hurts 2021-10-17 03:38:38 are all of your packages up to date? 2021-10-17 03:39:31 As of a few hours ago when I last checked, yeah. apk update && apk upgrade -al 2021-10-17 03:39:43 Let me re-test. I know everyone's been busy bees all day packaging stuff 2021-10-17 03:41:55 Yeah, fully current 2021-10-17 03:42:17 what architecture and CPU are you using? 2021-10-17 03:42:27 icu-libs-69.1-r0 2021-10-17 03:43:14 I see this: https://gitlab.alpinelinux.org/alpine/aports/-/issues/13086 2021-10-17 03:43:24 x86_64, Broxton/Silvermont Celeron N3450 2021-10-17 03:44:08 Yeah, that seems like what I 2021-10-17 03:44:13 Sorry, I'm getting 2021-10-17 03:45:16 I probably can't like, downgrade to the prior version without breaking who knows what, right? 2021-10-17 03:46:29 I;m not sure how downgrades work, but someone else here might know 2021-10-17 03:47:56 Called Bleeding-Edge for a reason, I guess 2021-10-17 03:48:02 Gonna get bit every once in a while 2021-10-17 04:36:49 Changed channels fine on my WSL2 instance. Gonna try on hardware now 2021-10-17 04:36:58 Hoping the APK blessings hold up haha 2021-10-17 04:51:02 Saijin_Naib[m]: i think you can just rebuild libffi locally and use that for now 2021-10-17 04:51:20 ABI should stay the same, so it should work 2021-10-17 04:51:26 at least that fixed the problem for me 2021-10-17 04:58:32 Is that a pretty straightforward abuild? 2021-10-17 04:59:57 I shudder to think how long that will be on this poor Celeron, haha 2021-10-17 05:00:24 So, clone repo, abuild - r, and install untrusted and away I go? 2021-10-17 05:00:40 Or which libffi do I need? 2021-10-17 05:05:39 In terms of mixing things across channels, getting Linux-LTS from Edge isn't too dangerous, right? I need my hardware enablement, haha. My MR has not landed in 3.14 2021-10-17 05:06:04 Userspace is all 3.14 now 2021-10-17 05:11:06 Bad idea or not, I'm up and running and I have hardware enablement to boot! Not too bad. Jeez... Alpine is really robust. No joke. I can't imagine having done this successfully under Debian or Sabayon or Solus 2021-10-17 05:11:43 And APK is so freaking fast that it was a matter of maybe 3 minutes to change channels and get everything sorted. Not some protracted like, 45min ordeal 2021-10-17 08:50:06 Qa: Is swi-prolog xpce only for graphics (So should it be in a separate package like Ubuntu does it?) 2021-10-17 11:46:09 Where can I find a jpg or png of the Alpine Linux logo? 2021-10-17 11:47:00 https://alpinelinux.org/alpinelinux-logo.svg 2021-10-17 11:47:10 or do you mean without the text 2021-10-17 12:07:45 Without the text 2021-10-17 12:12:25 then i only found this random site https://www.vectorlogo.zone/logos/alpinelinux/alpinelinux-icon.svg 2021-10-17 12:22:50 Oh lol. Is it not in a repo somewhere? 2021-10-17 12:23:43 https://gitlab.alpinelinux.org/alpine/infra/alpine-mksite/-/tree/master/_static 2021-10-17 12:24:03 ha, missed that one 2021-10-17 12:26:21 Ah yes there we go, thanks! 2021-10-17 12:27:06 open that svg in inkscape and remove the text yourself? ;) 2021-10-17 12:30:12 how is it with apk and purging modified configs etc created by the original install of the package? 2021-10-17 12:30:27 does "apk del " do this automatically or do I need some additional syntax ? 2021-10-17 12:30:58 thinking of debians "apt-get autoremove --purge" who removes unused stuff incl whatever leftovers there might be 2021-10-17 12:33:12 there is --purge as well 2021-10-17 12:38:10 How do I make /tmp to be tmpfs? Should I modify /etc/fstab or is there some other way (e.g. flag in openrc conf)? 2021-10-17 12:44:57 Gentoo suggests modifying /etc/fstab (https://wiki.gentoo.org/wiki/Tmpfs#OpenRC), I assume the same applies for Alpine? 2021-10-17 12:45:22 yes 2021-10-17 12:46:13 Thanks 2021-10-17 12:51:27 ikke: is it safe to use? 2021-10-17 12:51:38 and can --purge be runned later or only when del is used? 2021-10-17 12:51:53 Apachez: I think only when you remove the package 2021-10-17 12:52:14 darn... 2021-10-17 12:53:06 is there some way to check with installed packages? 2021-10-17 12:53:12 could audit be used for that? 2021-10-17 12:53:24 I mean to find out what can be purged or is unused or so 2021-10-17 12:54:33 Apachez: you can check /etc/apk/world for see the installed pacakges 2021-10-17 12:54:53 unused deps are automatcamilly removed 2021-10-17 12:54:59 youch, automatically* 2021-10-17 12:56:01 donoban: this is about config files 2021-10-17 12:56:33 ahh nice... 2021-10-17 12:56:40 that world to see whats installed 2021-10-17 12:56:40 ahm, ok 2021-10-17 12:56:51 but in my case its also about how to remove unused stuff to shrink the image 2021-10-17 12:56:58 but do this in a safe way 2021-10-17 12:57:04 similar to debians "apt-get autoremove --purge" 2021-10-17 12:57:45 with apk, autoremove is not needed 2021-10-17 12:58:06 yeah but I missed that purge thingy 2021-10-17 12:58:07 if you 'apk del something' the unused deps of something will be automically removed too 2021-10-17 12:58:17 but hopefully those leftovers aint that many bytes 2021-10-17 12:58:22 ah I see 2021-10-17 12:59:01 well maybe you can fin something using lbu 2021-10-17 12:59:39 I don't use it since years ago but I think that there is some comand that list you the files that differ from original apk 2021-10-17 12:59:57 if you have unused config files from deleted packages, it should list them 2021-10-17 13:02:08 that would be nice as a hint 2021-10-17 13:02:24 like match installed apks and their content with whats actually on disk which doesnt match that list 2021-10-17 13:02:25 maybe it just compare with last commit image instead the original apk files 2021-10-17 13:02:31 I'm not sure :S 2021-10-17 13:02:42 this is diskmode running as vm 2021-10-17 13:02:46 so no lbu 2021-10-17 13:03:46 maybe as you said what I mean is apk audit 2021-10-17 13:04:13 for example why is .X11-unix and .ICE-unix created upon boot in /tmp when this installation have no X11 running? 2021-10-17 13:04:57 bootmisc creates them 2021-10-17 13:05:32 it's just an empty dir on a tmpfs 2021-10-17 13:05:51 audit List changes to the file system from pristine package install state 2021-10-17 13:06:01 so A and U means? :P 2021-10-17 13:07:22 Added and Updated ? 2021-10-17 13:45:00 Apachez: best to check lbu source in such cases 2021-10-17 14:56:48 there is no lbu when you use diskbased install 2021-10-17 14:57:41 ala system disk mode 2021-10-17 16:47:34 oh, you mean apk audit 2021-10-17 17:17:39 Qa 2021-10-17 17:19:40 ktprogra1s, yes, xpce is only for graphical environment which include a nice prolog editor, thread monitoring, class inspection, etc. it is only for gui, so it could be put in a separate package 2021-10-17 17:20:32 ktprogra1s, (re: swi-prolog) 2021-10-17 17:59:21 ktprogra1s, you may want to put the java connectivity in another package also 2021-10-17 18:18:47 Did the CPU requirements for edge change? I'm getting Illegal Instruction from libffi in a variety of apps since updating this morning. 2021-10-17 18:18:57 ARF: nop 2021-10-17 18:19:06 ARF: try to upgrade libffi 2021-10-17 18:19:53 oh I guess I just have incredibly bad timing. 2021-10-17 18:20:00 thanks, ikke 2021-10-17 18:20:14 We pushed this fix about an hour ago 2021-10-17 18:20:43 I've been trying to figure out what happened for about that long, so it makes sense. 2021-10-17 19:51:35 Oooh, so Edge should be safe to move back to now? Hmm... 2021-10-17 19:51:44 When about is 3.15 expected to be pushed? 2021-10-17 19:52:05 Debating going back to Edge or staying safe on here so I can keep working in GNOME Builder without fuss 2021-10-17 19:52:11 Scheduled for beginning november, but depends on when the builders are finished 2021-10-17 19:53:06 Makes sense. Coming up fast. Which means edge is going to semi-stabilize shortly right, as all the major stuff has landed? 2021-10-17 19:53:32 Some boost rebuilds, and perhaps ruby as well 2021-10-17 19:53:55 Alright. Does boost affect things the same way libffi did? 2021-10-17 19:54:07 Mismatched builds vs provided boost making things segfault? 2021-10-17 19:54:28 Potentially it could 2021-10-17 19:54:54 Does Rust stuff work differently? None of my Rust stuff crapped out with the libffi changes 2021-10-17 19:55:23 static builds. They are probably still using the old libffi if it used it 2021-10-17 20:04:38 seems like busybox' wget only looks at http_proxy/https_proxy and not HTTP_PROXY/HTTPS_PROXY 2021-10-17 20:04:49 i'd consider that a bug, or is that well-defined behavior? 2021-10-17 20:11:15 https://httpoxy.org/ 2021-10-17 20:56:35 seems a bit strange - it still looks at environment variables for proxy settings, just under the lowercased name 2021-10-17 20:56:47 I don't see how it could be relevant to that 2021-10-17 20:57:32 aren;t the proxy environment settings always lowercase? 2021-10-17 20:57:39 I have at least only seen it like that 2021-10-17 21:02:41 everything i can think of respects both upper and lower case, but i believe i mostly see it upper cased in documentation 2021-10-17 21:03:29 i was wondering why wget failed to download something in an isolated network while everything else worked, i had to fiddle around for quite a bit to figure out that busybox' wget isn't like other wgets 2021-10-17 22:01:30 technically only http_proxy needs to be lowercase 2021-10-17 22:02:00 which is why i posted the link to explain it 2021-10-17 22:17:07 i see 2021-10-18 05:26:23 hi, I'm new here 2021-10-18 05:26:56 I need some help building sdcc on my alpine 3.14 2021-10-18 05:28:02 (maybe needless to say, but its build system is garbage) 2021-10-18 07:59:47 anyone in here who have successfully used librenms with alpine as os? doesnt seem to exist as a premade apk (yet) 2021-10-18 08:45:09 Qa: Can you please send your (can be anonymous) name and email so I can put you as the maintainer for the swi-prolog package? 2021-10-18 10:59:40 Anyone have recommended reading for how to debug issues with apps that crash when built on musl? I've been trying to get a number of ham radio apps to work (fldigi, wsjtx, hamlib, etc) 2021-10-18 11:20:25 ARF: https://wiki.musl-libc.org/functional-differences-from-glibc.html is a good place to start 2021-10-18 11:21:53 ARF: https://wiki.musl-libc.org/compatibility.html in general is useful -- other distributions may have had patches you can look into as well 2021-10-18 11:33:47 rcf: Thanks, I'll take a look there. 2021-10-19 07:54:32 how do i get the microphone working on alpine? 2021-10-19 07:55:14 firefox detects a microphone 2021-10-19 07:55:25 but doesn't get audio from it 2021-10-19 07:58:30 alsa? or 2021-10-19 07:58:57 if yes see https://bbs.archlinux.org/viewtopic.php?id=252378 2021-10-19 07:59:43 Hello, any ideas as to why i'd suddently (between reboots) be getting permission denied while accessing the seatd socket when starting sway? 2021-10-19 08:00:35 i'm using pulse 2021-10-19 08:00:39 needed it for bluetooth audio 2021-10-19 08:01:48 amk i have the same problem atm 2021-10-19 08:02:29 also, the microphone is detected 2021-10-19 08:02:36 just no audio comes out of it according to firefoc 2021-10-19 08:02:39 x 2021-10-19 08:02:53 hm 2021-10-19 08:03:10 firefox produces audio just fine 2021-10-19 08:04:15 Misthios: I was able to start sway just now by chowning /run/seatd.sock but this doesnt seem like a great solution, seatd does accept a user/group to set the socket owner so maybe editing /etc/init.d/seatd might be the solution 2021-10-19 08:04:35 yeah i do chown as well atm 2021-10-19 08:06:02 https://gitlab.alpinelinux.org/alpine/aports/-/commit/05b4a70abb1fd83434f1518ee0de36625e7d7fa2 this is the last commit from sway 2021-10-19 08:08:09 wack, i guess thats it. i wonder how to fix it properly 2021-10-19 08:09:17 huh, opening pavucontrol 2021-10-19 08:09:24 setting the audio stereo duplex 2021-10-19 08:09:27 to something else 2021-10-19 08:09:29 then back to it 2021-10-19 08:09:31 fixed it 2021-10-19 08:11:13 hm 2021-10-19 12:12:00 amk: Huh, in /etc/init.d/seatd group of seatd socket file is being set to 'seat' with command_args, works for me 2021-10-19 12:13:45 Are you in 'seat' group? You need to be in it to run sway via seatd daemon, but that could have been ignored when sway was suid 2021-10-19 12:47:41 dekedro[m]: Aha, this fixed it , thanks very much! 2021-10-19 16:59:27 thanx for fixin' that libffi issue 2021-10-19 18:38:55 ccx^xmpp: "The main reason I don't use netsurf..." You might be interested in visurf, which is netsurf with vi keybindings. 2021-10-19 18:42:09 Likely, thanks 2021-10-19 18:54:41 visurf? 2021-10-19 19:56:48 fun fact: Luke Smith's webpage isn't optimized for NetSurf 2021-10-19 20:10:19 strajder: in what way? 2021-10-19 20:11:09 elinks seems to work quite weel 2021-10-19 20:11:11 well 2021-10-19 20:11:35 did you open it in NetSurf? 2021-10-19 20:11:48 no i dont have it 2021-10-19 20:12:09 try it, you'll see; text is overlapping other text, colors are bonkers... 2021-10-19 20:13:05 damn 2021-10-19 20:13:14 probably it uses CSS3 features, when NetSurf uses CSS2.1 2021-10-19 20:13:23 sould netsurf work for netbank? 2021-10-19 20:13:51 *could 2021-10-19 20:14:42 he also has full articles in rss :] 2021-10-19 20:14:45 idk, but the general rule is: if it works with CSS 2.1 and no JS, then it will work in NetSurf 2021-10-19 20:15:48 mhmhm 2021-10-19 20:16:01 for example sr.ht works 2021-10-19 20:16:37 github is problematic, but has some basic functionality 2021-10-19 20:16:53 ahh :] 2021-10-19 20:18:39 netsurf seems like kinda a weird niche between modern browser and text only tho... 2021-10-19 20:19:33 it is blazing fast though; starts almost instantly 2021-10-19 20:19:48 ah thats cool 2021-10-19 20:20:07 elinks very fast too :p 2021-10-19 20:20:12 and i guess surf 2021-10-19 20:20:23 jaja.. 2021-10-19 21:53:23 Hi, 2021-10-19 21:53:23 /etc/rc.conf --> rc_cgroup_mode="unified" 2021-10-19 21:53:23 However Docker itself seems to work just fine. 2021-10-19 21:53:23 In my setup, Alpine 3.14.2 with cgroups V2 enabled, docker stats does not report MEM USAGE / LIMIT, MEM % and BLOCK I/O. 2021-10-19 21:53:23 Any ideas how to get the stats working for all metrics? 2021-10-19 22:11:53 does `docker info` have any warnings 2021-10-19 22:14:35 docker shows no warnings 2021-10-19 22:15:45 and it shows 'Cgroup Version: 2' 2021-10-20 01:36:23 Hi guys. Just wanted to ask, Does 1,0mesa-dri-intel99,99 support 2021-10-20 01:36:44 Supports Intel Mobile 945GM? 2021-10-20 01:39:08 Tried to install it before, but got black screen when starting sddm. 2021-10-20 01:39:38 the x startup log would be useful 2021-10-20 01:41:13 also did you install xf86-video-intel or just mesa-dri-intel 2021-10-20 01:42:57 I installed xf86-video-intel with mesa-dri-intel 2021-10-20 01:43:17 When SSH-ing, The X startup log does not show any error 2021-10-20 01:43:28 While SDDM just exit immediately 2021-10-20 01:43:45 And leaving the screen to be black and nothing 2021-10-20 01:45:32 In /var/log/sddm.log only says that they're exit without error or whatsoever 2021-10-20 01:46:21 When starting plasma-x11 without sddm, i got a error about segment fault for plasmashell 2021-10-20 01:46:55 That's what i can get. 2021-10-20 01:48:48 Launching X and glxinfo works well 2021-10-20 01:48:53 But not for plasma 2021-10-20 01:55:01 And no. k 2021-10-20 01:55:12 *And no my PC is not using NVIDIA 2021-10-20 17:53:25 Hi folks, I am working on https://gitlab.alpinelinux.org/alpine/aports/-/issues/13088#note_186740 2021-10-20 17:53:39 and trying to setup integration tests for Alpine 2021-10-20 17:54:02 I am trying to install Alpine to a chroot 2021-10-20 17:54:08 my current OS is Arch Linux 2021-10-20 17:54:21 So I run `sudo apk --root $mount --initdb --update-cache --allow-untrusted add alpine-base` 2021-10-20 17:54:39 it installs the packages successfully 2021-10-20 17:54:59 but when I try to boot it with QEMU and Arch Linux kernel it gets stuck: 2021-10-20 17:55:13 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/OKGGDRxPBSRppkDPGxiMpKJd 2021-10-20 17:55:15 alpine-base is not sufficient for booting 2021-10-20 17:55:45 what other packages are required to get to login prompt? 2021-10-20 17:55:55 q: why boot a chroot with qemu 2021-10-20 17:56:24 eris: these are integration tests for Alpine + booster initramfs 2021-10-20 17:56:32 If you want to boot it with qemu, I would use the virt image 2021-10-20 17:56:51 My tests use QEMU 2021-10-20 17:57:03 Here is an example for VoidLinux https://github.com/anatol/booster/blob/master/tests/generate_asset_voidlinux.sh 2021-10-20 17:57:33 these are not performance tests 2021-10-20 17:58:01 it more a simple functional test to make sure initramfs stage does not have any issues with user-space init like OpenRC 2021-10-20 17:59:25 So returning back to my question. How to debug Alpine's OpenRC and understand why it stuck? 2021-10-20 17:59:42 My guess some package are not installed or some services (like tty) are not started 2021-10-20 18:00:46 AnatolPomozov: if you want to use the result as a VM image then you'll need to install a kernel, bootloader, etc also 2021-10-20 18:01:00 does it get far enough that you can even see OpenRC loading things? 2021-10-20 18:01:26 If so, you can maybe try putting it into interactive mode by editing /etc/rc.conf 2021-10-20 18:01:53 so as ikke pointed out you would need to specific linux-virt package (or linux-lts) to get a kernel, and also either syslinux or grub for bootloader 2021-10-20 18:02:37 minimal: I specify kernel to QEMU directly, it does not require a bootloader. 2021-10-20 18:03:36 AnatolPomozov: ok, when I build QEMU images using apk in chroot I install the kernel & bootloader inside the resultant VM image 2021-10-20 18:04:19 if you don't install kernel inside image then what about any kernel modules that may be required to be loaded during boot? 2021-10-20 18:05:06 minimal: it depends on a way QEMU started. If `-kernel` parameter points to a Linux kernel then bootloader is not used. My other tests do it for Arch Linux and Void Linux 2021-10-20 18:05:44 minimal - good point about the kernel modules. They need to be a part of root fs. 2021-10-20 18:05:44 AnatolPomozov: also did you enable relevant init.d services when you built the VM image? 2021-10-20 18:06:50 minimal - I did not enable any services. I am trying to understand what need to be done. 2021-10-20 18:06:56 I looked at https://github.com/nmeum/alpine-chroot/blob/master/alpine-chroot-create as an example 2021-10-20 18:07:03 AnatolPomozov[m]: you can look at setup-alpine (and related scripts) to see what services are enabled 2021-10-20 18:07:32 that's likely you problem then - you could have a look at my script for creating Apline images :-) https://github.com/dermotbradley/create-alpine-disk-image 2021-10-20 18:07:46 it does exactly what you need (plus more) 2021-10-20 18:09:21 thanks minimal, will look at it. It sounds like The solution is here https://github.com/dermotbradley/create-alpine-disk-image/blob/c3c0cee67cc71bcd65def359e316c7ddcd3441e4/create-alpine-disk-image 2021-10-20 18:09:42 and btw how to enable debug mode for Alpine's OpenRC? 2021-10-20 18:11:18 AnatolPomozov: if you run "./create-alpine-disk-image --boottype bios --script-filename create.sh" and have a look at the resultant create.sh file that will show the sequence of commands to create a QEMU compatible image 2021-10-20 18:11:18 rc.conf `rc_verbose=yes` I guess 2021-10-20 18:12:13 thank you minimal - this script will be helpful to understand the installation process 2021-10-20 18:15:06 looking at create.sh, thank you @minimal 2021-10-20 19:08:07 AnatolPomozov[m]: the "standard" setup is https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimage.sh + https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/scripts/mkimg.standard.sh then https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in. alpine-conf (setup-*) doesn't do that much, it basically just copies /etc to a new 2021-10-20 19:08:09 partition 2021-10-20 19:08:28 s/setup-\*/setup-disk/ i guess 2021-10-20 19:09:12 see https://gitlab.alpinelinux.org/alpine/mkinitfs/-/blob/master/initramfs-init.in#L627 in particular 2021-10-20 20:34:09 hi 2021-10-20 20:56:05 Does anyone use powertop under linux-edge? 2021-10-20 21:55:13 Saijin_Naib[m]: it doesn't work because linux-edge is compiled without debugfs (or it misses something similar) 2021-10-20 21:55:51 donoban: Makes sense, and basically what the error indicates. 2021-10-20 21:55:52 also if you run int on a laptop take care that linux-edge runs peformance governor 2021-10-20 21:56:02 Wanted to A/B test power consumption of lts vs edge, but yeah 2021-10-20 21:56:18 I force the intel pstate powersave governor and use tlp 2021-10-20 21:56:29 ah great 2021-10-20 21:57:56 Yeah! edge seems nice. No stability issues vs lts now that mps added support for my hardware 2021-10-20 21:59:48 hehe nice 2021-10-20 22:00:02 I'm absolutely tired, time for sleep see you 2021-10-20 22:00:06 Peace! 2021-10-20 23:24:32 Does anyone know how to properly use musl-locales? I'm trying to get an app that chokes on the "C" locale to work (wsjtx) and the documentation is a bit sparse. 2021-10-21 18:19:24 Anyone feel like helping me sanity check this APKBUILD? I'm manually patching a makefile which is a first for me... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/iOjglqSVPLwziEXWJKFREDNP) 2021-10-21 18:20:31 Saijin_Naib[m]: I would advise using patch files 2021-10-21 18:20:48 Instead of what I'm doing here? 2021-10-21 18:20:48 They are easier to follow 2021-10-21 18:20:50 yes 2021-10-21 18:20:57 Hmm. Functionally the same thing, though? 2021-10-21 18:21:03 Like, I can use my sed lines? 2021-10-21 18:21:14 Those were a nightmare to write haha 2021-10-21 18:21:26 yes, it's possible to use them 2021-10-21 18:21:45 but they can unintentionally match more than you expect 2021-10-21 18:21:54 and harder to follow what the goal is 2021-10-21 18:22:26 Alright 2021-10-21 18:22:47 Is there a GUI way to go about making the patchfile? 2021-10-21 18:22:54 When upstream changes things, a patch won't apply, given you a signal that you need to check again 2021-10-21 18:23:02 Oh, that's cool 2021-10-21 18:23:12 Whereas sed will happily do whatever? 2021-10-21 18:23:22 check out the git repo, change the Makefile to what you want it to be 2021-10-21 18:23:23 yes, or do nothing 2021-10-21 18:23:38 then use `git diff > changes.patch` to get a patch file 2021-10-21 18:23:49 I typically run `git init -q && git add . && git commit -qm.` in the builddir 2021-10-21 18:23:57 mmm 2021-10-21 18:23:57 then make changes and use git diff to generate the patch 2021-10-21 18:24:01 I'm working out of tree 2021-10-21 18:24:12 Since I'm still struggling with clean commits and squashing when I make MRs 2021-10-21 18:24:21 I want to just push once I'm done 2021-10-21 18:24:32 it's unrelated, you can go clone it in /tmp, make the .patch, and copy it over 2021-10-21 18:24:46 it's easier than writing a patch by hand 2021-10-21 18:24:53 or use what I do 2021-10-21 18:25:01 create a repo in the builddir 2021-10-21 18:25:46 (some projects only provide a tarball) 2021-10-21 18:26:27 Hmm, okay, that makes sense 2021-10-21 18:26:39 if you want to keep a commit message in the patch, you can commit it onto the repo and then use `git format-patch -1 HEAD` too 2021-10-21 18:26:46 Let me see what I can figure out 2021-10-21 18:29:07 Oh, sweet, we have Meld 2021-10-21 18:29:11 I've used that before 2021-10-21 18:36:05 Do we have a $CXX variable for cxx=g++ like we do for CC=$CC 2021-10-21 18:36:42 https://gitlab.alpinelinux.org/alpine/abuild/-/blob/master/functions.sh.in#L163 2021-10-21 18:36:50 Or actually line 170 2021-10-21 18:37:08 Sweet, okay, thank you 2021-10-21 18:37:14 Trying to un-static this Makefile 2021-10-21 18:39:15 Alright, so where my Makefile says CC = GCC, I should be putting CC=${CROSS_COMPILE}gcc 2021-10-21 18:39:25 Or the $CC from HOSTCC =? 2021-10-21 18:39:42 you should put $CC 2021-10-21 18:39:57 Because I'm not cross-compiling? 2021-10-21 18:40:09 because $CC correctly sets the cross compiling itself if you are 2021-10-21 18:41:07 Oooh, okay 2021-10-21 18:41:10 Thank you for helping me understand 2021-10-21 18:45:39 Meld generated this patchfile:... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/hyCafcSMHDJklOjpdaUEpljz) 2021-10-21 18:45:54 looks sane 2021-10-21 18:46:56 Sweet 2021-10-21 18:47:06 oh, the path 2021-10-21 18:47:15 it should not be an absolute path 2021-10-21 18:47:17 The lines without the -/+, can those be dropped? 2021-10-21 18:47:33 No 2021-10-21 18:47:38 they are context lines and are important 2021-10-21 18:47:41 Yeah, that makes sense to make it more generalized. How would that look? 2021-10-21 18:47:53 oh... that's how it finds which "instance" of whatever is being modified? 2021-10-21 18:48:01 Like a big match block? 2021-10-21 18:48:23 --- a/Makefile, +++ b/Makefile 2021-10-21 18:48:31 That's what git would do 2021-10-21 18:49:24 Thank you! Adjusted 2021-10-21 18:49:27 let's run it 2021-10-21 18:49:33 Saijin_Naib[m]: if lines move, it can still find them 2021-10-21 18:49:33 Wait, I have to source it 2021-10-21 18:50:02 yes, add it to source, with a .patch extension 2021-10-21 18:50:20 Newline? Comma? Colon? 2021-10-21 18:50:29 How do I separate it from the existing tar get 2021-10-21 18:50:38 The docs just say add it haha 2021-10-21 18:50:44 I'm... below that familiarity 2021-10-21 18:51:21 typically one per line, no delimiters except whitespace 2021-10-21 18:52:11 check how existing packages do it 2021-10-21 18:52:35 Yeah, I'm digging to find one that has it similar 2021-10-21 18:53:10 This one is also weird since it has a cmd, gui, and docs 2021-10-21 18:53:24 and they should all be split out 2021-10-21 18:54:28 this is an example I was just looking at https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/26606/diffs 2021-10-21 18:56:15 Thank you 2021-10-21 18:58:59 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/vLAipxGekwCiMePRCQflYAZQ 2021-10-21 19:09:36 your patch is wrong 2021-10-21 19:09:50 could you post it somewhere 2021-10-21 19:51:06 https://gist.github.com/Saijin-Naib 2021-10-21 19:51:15 Thanks for helping, as ever, psykose 2021-10-21 19:53:24 hmm 2021-10-21 19:54:57 I'm sure my patching the Makefile broke it somehow 2021-10-21 19:55:10 It's a pretty convoluted Makefile 2021-10-21 19:55:21 From all of my like, two weeks experience 🤣 2021-10-21 19:57:24 sorry, make your patch delete the lines instead 2021-10-21 20:00:38 the default $CC is `cc`, which is a symlink to the default compiler, gcc in this case, ditto for $CXX 2021-10-21 20:00:56 however if you set CC=$CC, what happens is it needs to be passed like `make CC=something all`, otherwise it's blank 2021-10-21 20:02:04 the default $CC can read env vars, eg `CC=something make` which get set in abuild, but not if it's overriden with CC=$CC 2021-10-21 20:02:07 my mistake :) 2021-10-21 20:02:28 Ooh okay. 2021-10-21 20:02:36 So, CC, GXX, 2021-10-21 20:02:41 What about the $xFLAGS ones? 2021-10-21 20:02:42 Same deal? 2021-10-21 20:03:53 remove cflags, rest should be fine i think 2021-10-21 20:04:11 CC=$CC is fine except it's wrong syntax, should be CC=$(CC) for make format. but it is redundant 2021-10-21 20:04:16 ditto for ldflags 2021-10-21 20:04:23 i still don't understand why you cannot just do make CC=gcc 2021-10-21 20:04:32 or, well, make CC=$CC 2021-10-21 20:04:56 make CC=$CC CXX=$CXX CFLAGS="$CFLAGS" LDFLAGS="$LDFLAGS" prefix=/usr 2021-10-21 20:05:38 CC=$(CC) is recursive, you would need := to do that, but yes it's basically just the default rule 2021-10-21 20:06:07 ah, yes, := 2021-10-21 20:07:04 Alright, so nuke most everything in my patch except the adjustment of the prefix directory 2021-10-21 20:08:13 even for prefix you can do prefix= on the make invocation, so you can drop the patch entirely if you want 2021-10-21 20:08:38 It won't build unpatched 2021-10-21 20:08:47 Unless my APKBUILD is also malformed 2021-10-21 20:08:59 Oh nevermind 2021-10-21 20:09:04 I misunderstood what you wrote 2021-10-21 20:09:07 I have not tried that 2021-10-21 20:12:03 https://gist.github.com/Saijin-Naib/ffb2049bea3257eade9bd92b69af6fdc 2021-10-21 20:12:15 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/MuZsSLLGfBWVzRBSIsYUEdeF 2021-10-21 20:13:14 Commenting out the subpackage step leads to what appears to be a successful build 2021-10-21 20:13:17 the -doc subpackage function didn't make anything, probably because your prefix is set wrong 2021-10-21 20:13:30 Even with the prefix= thing? 2021-10-21 20:13:51 So I may need to patch it? 2021-10-21 20:15:20 you have to do prefix=/usr 2021-10-21 20:16:33 Ahhh 2021-10-21 20:19:18 Wow, built perfectly, thank you! 2021-10-21 20:19:21 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/dCRfzhqNzGfOZPfxspxvBRpR 2021-10-21 20:19:26 Doesn't run in GUI mode, however, haha 2021-10-21 20:23:45 it uses gtk2 and you have gtk+3.0 2021-10-21 20:25:18 https://github.com/dfandrich/gpscorrelate/blob/master/INSTALL#L23 2021-10-21 20:25:36 Which I hope means it is meant for GTK3 2021-10-21 20:25:52 Because otherwise, they're just making a trap haha 2021-10-21 20:26:32 building with only gtk+3.0-dev gives warnings that pkgconf cannot find gtk2 2021-10-21 20:26:35 and the makefile says as much 2021-10-21 20:26:54 there is only gtk+2.0 in the makefile, unless i am missing something 2021-10-21 20:27:21 unless.. 2021-10-21 20:28:16 ah 2021-10-21 20:28:33 i cloned the wrong repository 2021-10-21 20:28:34 of course 2021-10-21 20:29:15 somehow i manage to do this every time for forked software 2021-10-21 20:30:39 Oh yeah, I have no idea what upstream's state is 2021-10-21 20:30:44 This is the one that is currently flatpak packaged 2021-10-21 20:30:47 That's all I know 2021-10-21 20:32:57 it's trying to load a config file with the g_key_file* functions 2021-10-21 20:33:48 That sounds like something it should be providing 2021-10-21 20:34:41 it fails to load the file because it doesn't exist, then there is a sigsegv because... they reuse an error object and the library doesn't like that 2021-10-21 20:34:43 or whatever 2021-10-21 20:34:58 anyway the issue here is they don't handle errors correctly, but if you make some correct config file somewhere it should work 2021-10-21 20:35:11 Hmm 2021-10-21 20:35:19 I wonder how I can get that 2021-10-21 20:35:27 allow me to peruse more 2021-10-21 20:35:29 Can one like... download a flatpak and poke into it? 2021-10-21 20:36:17 technicallyt 2021-10-21 20:36:42 It would have to have that file, right? 2021-10-21 20:36:48 Or do we not have it because of MUSL or something 2021-10-21 20:37:17 no, it's the code at fault 2021-10-21 20:37:35 Oh, okay 2021-10-21 20:37:36 Thank you 2021-10-21 20:37:42 also funny, they use g_get_user_config_dir which get's XDG_CONFIG_HOME, but they put it in a variable called UserHomeDir 2021-10-21 20:38:19 Why would they do that? 2021-10-21 20:38:24 More "portable"? 2021-10-21 20:38:44 They make a note of using pure C99/POSIX/XOpen 2021-10-21 20:38:45 they try to load config_home/.gpscorrelaterc 2021-10-21 20:39:04 XDG_CONFIG_HOME by default should be `~/.config` 2021-10-21 20:39:13 UserHomeDir implies `~` 2021-10-21 20:39:50 Ahh 2021-10-21 20:43:51 i changed one line and it works 2021-10-21 20:44:48 Whaaat 2021-10-21 20:44:49 https://termbin.com/2dwt 2021-10-21 20:45:05 No way 2021-10-21 20:45:07 Haha 2021-10-21 20:45:15 How do you folks figure this shit out. Beyond amazing to me 2021-10-21 20:45:26 maybe making ~/.config/.gpscorrelaterc also helps 2021-10-21 20:45:36 Saijin_Naib[m]: reading and trying 2021-10-21 20:45:36 nope, works without 2021-10-21 20:45:45 lots of experimenting and failure :) 2021-10-21 20:45:56 the library just sigsegvs if GError isn't null when trying to be written to 2021-10-21 20:45:59 a tiny bit of stubborness 2021-10-21 20:46:15 i didn't think not initialising it at first made it not null :p 2021-10-21 20:46:30 in my case, lots of stubborness 2021-10-21 20:46:37 feel free to use that as a patch if it works for you 2021-10-21 20:46:40 it's already in patch format 2021-10-21 20:46:42 day 5 of getting dotnet to build 2021-10-21 20:46:51 heh 2021-10-21 20:46:55 i don't envy you 2021-10-21 20:47:06 last i tried to build dotnet their auto script failed every time 2021-10-21 20:47:08 it builds, just not as alpine package 2021-10-21 20:47:41 psykose: Is that what the SEGFAULT was? Something was already loaded as garbage where it was supposed to be NULL? 2021-10-21 20:47:52 and rootbld makes cmake go `-j1` for some reason 2021-10-21 20:48:11 psykose: I am trying now, and I absolutely will, thank you so much 2021-10-21 20:48:15 i think the library refuses to overwrite non-null errors 2021-10-21 20:48:26 so the default gerror was garbage, then the code does error && error->something 2021-10-21 20:48:34 and ->something is a sigsegv on garbage 2021-10-21 20:48:42 but the && passes cause it's not null/0 2021-10-21 20:48:53 that's why you get that glib warning 2021-10-21 20:49:05 but of course it gets accessed while being invalid in the next line 2021-10-21 20:49:19 running the code in gdb/lldb with CFLAGS=-Og gives the specific line 2021-10-21 20:49:58 It runs! Oh, that's so freeaking awesome! Thank you! 2021-10-21 20:50:05 it's yet another case of uninitalised variables breaking someones code, i think the glibc malloc null initialises it 2021-10-21 20:50:09 so nobody else cuaght this 2021-10-21 20:50:18 What would I owe you for help to get the $pkgname-lang built, and split out to gui/cmd/doc/lang properly? 2021-10-21 20:50:22 or, well 2021-10-21 20:50:25 it's stack initialised 2021-10-21 20:50:28 but whatever happens there 2021-10-21 20:50:36 So... Would this be something I should upstream to them as a possible patch? 2021-10-21 20:50:44 i think so 2021-10-21 20:50:47 Or is this MUSL LIBC only because MUSL is more stringent (from my understanding)? 2021-10-21 20:51:15 So under GLIBC, it doesn't crash, but isn't that memory still garbage? Isn't that bad still? 2021-10-21 20:51:20 this patch won't break anything 2021-10-21 20:51:27 Saijin_Naib[m]: fyi, it's just musl 2021-10-21 20:51:32 there already are makefile functions for gui 2021-10-21 20:51:38 and docs 2021-10-21 20:52:06 make gpscorrelate-gui builds only the gui 2021-10-21 20:52:06 etc 2021-10-21 20:52:11 Yeah 2021-10-21 20:52:14 just call those instead of just make all 2021-10-21 20:52:16 So I need to make subpackages that call those, right? 2021-10-21 20:52:24 something like that 2021-10-21 20:52:26 as separate build lines 2021-10-21 20:52:29 Okay 2021-10-21 20:52:56 And the stuff for lang, does that require a patch to enable, or is that also something that can be invoked with the make command? 2021-10-21 20:55:12 How and where do I credit you for the patch? In the patch? The APKBUILD? Not a thing that's normally done? 2021-10-21 20:55:25 looks like build/install-po to me 2021-10-21 20:55:33 but there is no lang for this 2021-10-21 20:55:38 those look like translated docs 2021-10-21 20:55:41 in the po there are a few, I think 2021-10-21 20:55:43 Oh 2021-10-21 20:55:51 lang is for the program being translated 2021-10-21 20:55:57 these install under doc by default 2021-10-21 20:56:17 and the install location is correct, so you only need to declare gpscorrelate-doc as a subpackage, it should just work 2021-10-21 20:56:21 Ohh... I thought the po folder was for UI translation 2021-10-21 20:56:31 maybe it is actually, in which case the location is wrong 2021-10-21 20:56:32 let me see 2021-10-21 20:56:37 Yeap, that works fine. It breaks when I try to add the gpscorrelate-lang package 2021-10-21 20:57:19 lang is normally /usr/share/locale 2021-10-21 20:58:31 yeah it's just docs 2021-10-21 20:58:44 you can serve the folder it makes and visit it, it's just the same webpage as you find online 2021-10-21 20:58:46 and in french 2021-10-21 20:58:59 no -lang for this one 2021-10-21 21:00:41 Ah, crap. Their docs lied to me haha 2021-10-21 21:00:43 Thank you 2021-10-21 21:01:37 or maybe there is 2021-10-21 21:01:46 calling the gui package gpscorrelate-gui would be acceptable right? I see a few other packages have a similar naming schema 2021-10-21 21:01:57 Or would it be better to be gpscorrelate-gtk since it is GTK-based? 2021-10-21 21:03:08 -gui 2021-10-21 21:04:43 the fr.po does seem to have data for things inside the code, but the install function doesn't seem to install anything related to it 2021-10-21 21:05:34 Mabye incomplete? 2021-10-21 21:05:42 i wouldn't bother with it if it doesn't just work 2021-10-21 21:05:48 That's fair 2021-10-21 21:05:59 I just wanted to make sure I wasn't leaving out other lanaguages other folks could use if it was there 2021-10-21 21:06:06 But it appears it isn't there, so, I'm not hurting anyone 2021-10-21 21:06:54 nice to know it builds at least 2021-10-21 21:11:22 also funnily i put alpine on my main computer today, and all the flatpaks i used open in 1/10th the time 2021-10-21 21:16:02 win list 2021-10-21 21:16:13 Faster than native!? 2021-10-21 21:16:34 I was going to say, they're noticably heavy on my install, but not as bad as snaps were 2021-10-21 21:17:26 they were also flatpaks before, just now they open at native speed 2021-10-21 21:17:28 no idea why 2021-10-21 21:18:01 they were noticably heavy before indeed, opening skype/signal took like 4 seconds before a window showed 2021-10-21 21:18:28 i wonder how much of that is a casual move to alpine and how much is the fact i upgraded to an nvme drive 2021-10-21 21:18:37 i wouldn't think it is such a difference over a sata ssd 2021-10-21 21:18:51 or maybe the samsung 980 is just magic 2021-10-21 21:19:09 Probably the NVMe drive. 2021-10-21 21:19:36 The interface isn't the only thing you changed. Even in SATA mode the new drive probably performs better. 2021-10-21 21:19:37 the amount of files to load is really small though, i don't think sata is such a bottleneck for like 120MiB that is also in cache 2021-10-21 21:19:55 certainly not for... 4 seconds of difference 2021-10-21 21:20:01 Haha, that's massive 2021-10-21 21:20:04 ah well, magic it is 2021-10-21 21:20:10 Nothing else changed? lvm? underlying FS? 2021-10-21 21:20:23 I'm lvm+ext4 on my eMMC 2021-10-21 21:21:05 nope 2021-10-21 21:21:11 same old xfs and a fresh install 2021-10-21 21:21:21 and gentoo->alpine 2021-10-21 21:22:05 I don't think I've ever used Alpine on fast hardware, but it certainly feels fast on my ThinkPad T60... 2021-10-21 21:23:02 it definitely feels fast :p 2021-10-21 21:31:34 Fastest I've used it on is my desktop, so i7-6700k, 32GB DDR4, GTX 1050TI OC, and Crucial MX500 SSD 2021-10-21 21:31:37 It freaking flew 2021-10-21 21:31:58 Wonderful desktop OS 2021-10-21 21:32:10 A bit crap about nVidia not making a MUSL build of their drivers, but ah well 2021-10-21 21:33:22 No that's not the problem. 2021-10-21 21:33:50 big crap about nvidia not providing specs for their hardware to their own customers. 2021-10-21 21:34:09 Would that allow for the pstate thing to work properly in noveau then? 2021-10-21 21:34:44 It'd certainly have a better chance of happening. Maybe ask the nouveau devs? 2021-10-21 21:35:13 I'll never buy another nvidia card. The AMD drivers are great 2021-10-21 21:36:22 anyone use webkit2gtk browser for everything? just seems to break too often imo, surf/vimb/midori (btw elinks is super nice) 2021-10-21 21:37:42 eg was having big trouble viewing a tweet :/ 2021-10-21 21:38:01 Is epiphany that browser? 2021-10-21 21:38:08 If so, then yeah, it is touchy as hell 2021-10-21 21:38:25 elinks is amazing. Works great for when my DE/DM is broken and I need help haha 2021-10-21 21:38:35 Look stuff up in the console session until I sort it out 2021-10-21 21:38:41 waiting for NVMe heatsinks to arrive tomorrow so I can install Alpine on PC 2021-10-21 21:39:41 Do it! haha 2021-10-21 21:39:51 Do most people not daily it as a standard desktop install? 2021-10-21 21:39:53 trying to use it as much as possible for everyday stuff now :] 2021-10-21 21:40:17 It's been my daily on my laptops for almost two years now 2021-10-21 21:40:22 I mean, I already have alpine spanned on 2 SSDs 2021-10-21 21:40:26 Desktop is going to be a harder leap... I need my GPU 2021-10-21 21:40:36 I just never boot it into main desktop, it's always a VM for me 2021-10-21 21:42:42 oh, wow. That's a highly specific workflow. 2021-10-21 21:42:43 What hosts? 2021-10-21 21:43:18 Windows 2021-10-21 21:43:30 Ahhh... Hyper-V? 2021-10-21 21:44:03 Hyper-V, VMware 2021-10-21 21:44:53 depends what I'm doing, for building stuff it runs in Hyper-V since I access it via SSH, if I need desktop I use VMWare 2021-10-21 21:46:13 seems epiphany is webkit2gtk too yeah 2021-10-21 21:46:17 cool u using elinks haha 2021-10-21 21:46:33 elinks is a life-saver when you're terminal-only 2021-10-21 21:46:48 Download files, browse semi-comfortably... 2021-10-21 21:47:18 Yeah, Epiphany seems to load pages properly on the 2nd or 3rd refresh, but tends to bomb to an error screen on the first connection, like, 50% of the time 2021-10-21 21:47:45 Really odd browser, which sucks, because I liked that it offered Firefox Sync with Webkit engine and really good GNOME/GTK integration 2021-10-21 21:49:14 That summarizes my feelings about Epiphany too. It seems promising and I want to like it, but it just doesn't work some of the time. 2021-10-21 21:49:20 is there even a point in having them in repo?? 2021-10-21 21:50:47 I am hoping this is just temporary teething pains with it 2021-10-21 21:50:57 It is usable, but just touchy, I guess 2021-10-21 21:51:02 was thinking about downsizing and getting rid of qutebrowser/python, but i think ill stick to it 2021-10-21 21:51:11 It's great that it integrates the browser keyring with system, so you can use Seahorse to get at it 2021-10-21 21:51:32 never trusted keyrings :p 2021-10-21 21:51:33 I'm using Edge in flatpak right now 2021-10-21 21:51:39 I don't trust my memory 😛 2021-10-21 21:51:52 (without knowing much, also dont need/miss it much) 2021-10-21 21:52:26 Edge is on flathub? 2021-10-21 21:52:29 mhmm 2021-10-21 21:57:26 Hi guys, what is the best way to ask a question RE:alpine-linux? 2021-10-21 21:58:24 just ask a question? 2021-10-21 21:58:33 here 2021-10-21 21:58:50 OK thanks 2021-10-21 22:00:58 roughly I just installed alpine on a second partition along WindowsXP and syslinux bootloader can't offer me an option to boot on alpine 2021-10-21 22:01:19 Does it show anything? 2021-10-21 22:01:35 no just go directly on XP 2021-10-21 22:02:33 I added to extlinux.cfg to /boot/sda2 2021-10-21 22:03:22 but everytime I update update-extlinux removes the /boot part 2021-10-21 22:03:50 woof 2021-10-21 22:04:07 That's beyond me. I have not multibooted since like... ages. OS/2, SkyOS, etc 2021-10-21 22:04:16 And I used AirBoot to handle it all 2021-10-21 22:04:53 So, your 'XP' partition is set to active, then, it seems 2021-10-21 22:05:09 what partition flags do you have where your syslinux is? 2021-10-21 22:06:10 correct 2021-10-21 22:07:29 the XP was bootable and after I realized that I flag sda2 bootable but did not boot 2021-10-21 22:07:56 also I tried to install grub, all OK but not boot. 2021-10-21 22:09:40 mmm 2021-10-21 22:09:43 And grub was also marked active? 2021-10-21 22:09:52 Was it its own partition, or inside the MBR? 2021-10-21 22:10:02 Some BIOSs can do some weird stuff with MBR protection 2021-10-21 22:11:01 I am not sure what marked active means? 2021-10-21 22:11:58 what is /boot/sda2 2021-10-21 22:12:22 hold on a sec 2021-10-21 22:15:46 Well with MBR layouts, you can have one active partition that gets started at boot, and up to three other primary partitions 2021-10-21 22:16:16 If you have a dedicated partition for your bootloader like grub, then you want that marked active 2021-10-21 22:17:10 I see 2021-10-21 22:17:39 I tried update-grub but I think did not work 2021-10-21 22:17:57 You'd likely have to use something like parted/gparted to set the flags 2021-10-21 22:18:09 (or kparted, if you're under KDE) 2021-10-21 22:18:14 I followed some of this: https://wiki.alpinelinux.org/wiki/Installing_Alpine_on_HDD_dualbooting 2021-10-21 22:19:45 I used cfdisk ir fdisk 2021-10-21 22:19:51 I used cfdisk or fdisk 2021-10-21 22:31:24 Tip: If you use extlinux (default) and didn't create a separate boot partition, (/boot resides on the root partition /). 2021-10-21 22:31:27 You might need to adjust the paths to the boot files in /boot/extlinux.conf, eg.: 2021-10-21 22:31:30 [...] 2021-10-21 22:31:33 LABEL hardened MENU DEFAULT MENU LABEL Linux hardened LINUX /boot/vmlinuz-lts INITRD /boot/initramfs-lts 2021-10-21 22:39:43 @hello71: my bad, there is no /boot/sda2, I wanted to say LINUX /boot/vmlinuz-lts 2021-10-21 22:39:43 INITRD /boot/initramfs-lts 2021-10-21 22:40:05 in /boot/extlinux.conf 2021-10-21 22:41:06 I remember I update-extlinux and it removes /boot part or any changes, why? 2021-10-21 22:41:27 changes should be in some /etc/ file for syslinux 2021-10-21 22:41:38 it uses that to generate the /boot file 2021-10-21 22:51:08 yes, I found something on /etc and tested on my VMbox 2021-10-21 22:56:56 "/etc/update-extlinux.conf" 2021-10-21 23:05:43 we have to leave it there ... 2021-10-22 00:16:30 psykose https://github.com/dfandrich/gpscorrelate/commit/22fab45e43087f00e386e7d0e3327f5ce5efe3a4 2021-10-22 00:16:51 didn't need to credit me :) 2021-10-22 00:17:06 I never would have figured out that 🤣 2021-10-22 00:17:24 They moved fast! I can target that release and drop the patch now, right? 2021-10-22 00:17:28 i think this is the first commit in anything ever that has anything resembling me in it 2021-10-22 00:17:37 it makes more sense to have a patch on a release 2021-10-22 00:18:18 there are some amount of fixes since the last release (2019), you could ask for a 2.1 2021-10-22 00:19:03 unless something.. changes a ton, and you want all the fixes since a last release, you usually just take latest release (2.0) and patch specific things (the patch you have) on top of that 2021-10-22 00:22:26 some packages kinda stop making releases but have hundreds of commits since, in that case you could do lastreleaseversion_gitdate and just use something off master, but that's rare for things that have had any releases at all, usually just for 0_gitdate things that don't have releases 2021-10-22 00:23:10 in this case do nothing, just the patch, ask for new tag 2021-10-22 00:28:32 Sounds great. I'll ask. They seem nice 2021-10-22 00:38:04 whois 2021-10-22 01:35:22 Why is it that when I install less, it gets installed as /usr/bin/less.apk-new but /usr/bin/less is still a symlink to busybox? Is it safe to delete the symlink and rename the installed less? 2021-10-22 04:28:00 ktprograms: that's not normal 2021-10-22 07:54:03 ikke: Sorry, was busy and didn't see your msg till now. Is there any debugging info I can provide? 2021-10-22 10:27:06 have to say, I'm impressed, installed plasma and sway, did restart and everything just works, including listening to music over bluetooth to my soundbar 2021-10-22 11:19:44 Hello, Is there anything special I need to do in order to use a thunderbolt 3 dock? It doesnt seem to register at all that i've plugged it in -- no keyboard/mouse/monitors 2021-10-22 11:23:37 amk maybe firmware? Do you have usbutils and pciutils so you can lsusb and lspci and then sudo dmesg | grep firmware 2021-10-22 11:26:29 well I have no idea what changed but I just unpluged it and plugged it back in for the 50th time and it started working 2021-10-22 11:39:47 Gotta love computers 2021-10-22 11:40:11 Deterministic non-determinism 2021-10-22 12:03:47 Saijin_Naib: do I remember right that you asked about the possibility of enabling gnome-builder? 2021-10-22 12:04:04 oh yes, found it 2021-10-22 12:04:19 just wanted to send this: https://pkgs.alpinelinux.org/packages?name=gnome-builder&branch=edge 2021-10-22 14:02:22 does anyone know how i can search which package in repo contains a certain file? 2021-10-22 14:03:07 https://pkgs.alpinelinux.org/contents 2021-10-22 14:04:28 and in a terminal? 2021-10-22 14:07:32 apk info -W, if it's installed, otherwise idk 2021-10-22 14:09:14 back to searching i go 2021-10-22 14:39:17 Newbyte: Oh, thank you so, so, so much! If it stayed missing, I was going to try and stuggle-learn through re-enabling it 2021-10-22 14:40:07 I use it for technical documentation writing as it is seemingly the only IDE with first-class support for reStructuredText, and it is really wonderful to have that help from an IDE, as well as regular git workflow stuff in one spot. Much easier for someone like me to manage 2021-10-22 14:40:37 Was it hard to re-enable and update? 2021-10-22 14:41:37 Saijin_Naib: Not at all, the blocker had been fixed upstream 2021-10-22 14:42:10 Excellent! I hope it stays low-maintenance! I truly appreciate having it 2021-10-22 14:42:30 I think I'll try and tackle flatseal next 2021-10-22 14:42:52 Shouldn't be hard seeing the reason it was removed 2021-10-22 14:42:57 My flatpak one won't actually save permission changes for other flatpaks for whatever reason, and I can't find anything online about that 2021-10-22 14:43:31 That's reassuring at least. I'm hoping it was just disabled because nobody wanted it or something 2021-10-22 16:59:59 ERROR: unable to select packages: masked in: --no-network 2021-10-22 17:00:02 What does this mean? 2021-10-22 17:01:05 What command did you execute 2021-10-22 17:01:43 apk add coreboot-ectool (a package I made) 2021-10-22 17:02:20 what's in your /etc/apk/world and /etc/apk/repositories? 2021-10-22 17:02:32 Ah sorry, my package index wasn't signed 2021-10-22 17:02:34 Now it says masked in: @commie 2021-10-22 17:02:48 Not sure how being a commie has to do with anything 2021-10-22 17:03:02 probably to do with what's in /etc/apk/repositories 2021-10-22 17:03:39 Oh, the line with my repo had a @commie in front of it. 2021-10-22 17:03:45 I didn't do that though 2021-10-22 17:04:02 it's not something anything in alpinelinux does 2021-10-22 17:04:58 :# 2021-10-22 17:10:15 Seems like something somehow echo'd to that 2021-10-22 17:10:57 your alter ego edited the files 2021-10-22 17:11:46 Saijin_Naib[m], not even echo'd, it edited an existing line 2021-10-22 17:39:53 anyone else seeing that gnome-remote-desktop is kinda... empty? 2021-10-22 17:39:53 https://pkgs.alpinelinux.org/contents?branch=edge&name=gnome-remote-desktop&arch=x86_64&repo=community 2021-10-22 17:40:56 what is empty there 2021-10-22 17:41:10 no executable and no .desktop? 2021-10-22 17:41:31 isn't /usr/libexec/.. the executable 2021-10-22 17:42:06 gah 2021-10-22 17:42:06 damnit 2021-10-22 17:42:12 the desktop would be missing then 2021-10-22 17:42:14 this is the server for GNOME DE 2021-10-22 17:42:20 also all of those locale fines should be in -lang 2021-10-22 17:42:29 I think they changed the rmote desktop client to gnome connections 2021-10-22 17:42:35 Right 2021-10-22 17:44:14 Newbyte: Does Builder run for you? It refuses to launch but doesn't throw any errors in console 2021-10-22 17:44:47 Saijin_Naib: Yeah, it worked for me when I tested it 2021-10-22 17:44:48 I got the GUI open and all 2021-10-22 17:45:08 clang12 libs, too? 2021-10-22 17:45:42 Are you asking if I had clang12 installed? 2021-10-22 17:45:49 Yeah 2021-10-22 17:45:58 I just got upgraded to that from clang11 2021-10-22 17:46:27 builder opens fine for me 2021-10-22 17:46:28 I'm not sure 2021-10-22 17:46:54 psykose: just so we're on the same page - gnome-builder, right? 2021-10-22 17:46:58 yep 2021-10-22 17:47:07 Gah, super weird 2021-10-22 17:47:09 Darn... 2021-10-22 17:47:22 Do you have elogind running? 2021-10-22 17:47:24 i do 2021-10-22 17:47:24 Some apps needed that 2021-10-22 17:47:43 gnome-builder -v and also checking var/log/messages might have something 2021-10-22 17:47:52 Yep, in default runlevel 2021-10-22 17:48:30 add/del fixed it 2021-10-22 17:48:31 I'm good 2021-10-22 17:48:34 Thank you so much! 2021-10-22 17:49:26 did you by any chance have an error when adding it you missed 2021-10-22 17:49:33 because i did, needed to update first :p 2021-10-22 17:49:48 I had it orphaned from 3.14 install so it upgraded 2021-10-22 17:49:53 I think something odd happened because of that 2021-10-22 17:50:03 it's easy to miss any errors too 2021-10-22 17:50:06 I went 3.14 stable to get that and flatseal, then edge'd, then upgrade since then 2021-10-22 17:50:16 Yeah, entirely possible I ignored something 2021-10-22 17:50:24 I was worried it was me switching to clang12 libs 2021-10-22 17:50:45 Newbyte: haha, I know, bad idea, but I needed them both 😛 2021-10-22 17:51:21 both what 2021-10-22 17:51:31 flatseal and gnome-builder 2021-10-22 17:51:42 ah 2021-10-22 17:51:48 flatseal so I can give the flatpaks access to run/user/1000/gvfs/ 2021-10-22 17:51:49 why not just send an MR to revive Flatseal? 2021-10-22 17:52:02 you can do that without flatseal 2021-10-22 17:52:07 it just runs the flatpak commands for you 2021-10-22 17:52:08 I think I will 🙂 Just gotta study it 2021-10-22 17:52:23 override --filesystem=/run/user/1000/gvfs:ro or something 2021-10-22 17:52:27 i forget now 2021-10-22 17:52:40 git revert d4dafb44fbf0cf5e1662c4879379c4aebef53e90 and put yourself as maintainer 2021-10-22 17:52:43 then MR 2021-10-22 18:05:53 I'd do that from terminal then with that command? 2021-10-22 18:06:12 Looks like we can bump to release v1.7.4 as well if I take over 2021-10-22 18:15:34 Yeah 2021-10-22 18:15:41 In the aports directory 2021-10-22 18:15:47 How familiar are you with git? 2021-10-22 18:16:03 Enough to do damage 2021-10-22 18:16:09 Given a few of my prior MRs were a mess people had to correct for me 2021-10-22 18:16:11 That would be very little knowledge 2021-10-22 18:16:17 I prefer using the WebUI 2021-10-22 18:16:29 You won't once you learn how to use git 2021-10-22 18:16:43 But it takes a while to leran 2021-10-22 18:16:49 learn* 2021-10-22 18:17:06 Yeah, msotly what seems to be biting me is squashing commits and switching branches and stuff 2021-10-22 18:17:21 I can do pull, push, commit, etc 2021-10-22 18:17:23 Basic basic 2021-10-22 18:17:50 git reset and rebase are very useful 2021-10-22 18:18:10 but be careful with reset. very easy to delete all your work 2021-10-22 18:18:39 already done that haha 2021-10-22 18:18:47 I killed my catfish APKBUILD with that 2021-10-22 18:18:54 into the ether somewhere 2021-10-22 18:19:18 It's fine I guess since catfish needed some other dependencies if I recall, but still haha 2021-10-22 18:36:02 "enough to do damage" -- points for honesty 2021-10-22 19:05:50 invoked: Was fun waking up to see a bunch of people looking at my MR and going 😐️ 2021-10-22 19:05:54 So yeah, damage, haha 2021-10-22 19:07:06 fortunately it's very hard to break things in a reviewed merge request :) 2021-10-22 19:07:24 Yes, mostly just what little goodwill I have haha 2021-10-22 19:07:26 😛 2021-10-22 19:07:33 Making people work more than they need to already 2021-10-22 19:27:21 Newbyte: I leave Cogitri as the Contributor, right? 2021-10-22 19:27:21 Had to adjust the files a bit as libhandy1 is required now, and appstream-glib is required for runtime too 2021-10-22 19:27:45 Saijin_Naib: yes, keep that. just change maintainer 2021-10-22 19:27:56 and version and dependencies as necessary 2021-10-22 19:28:10 Alright, did the revert as you suggested, going to try and push it now. Hopefully it is just one commit 😐️ 2021-10-22 19:28:10 Yep, all adjusted, compiled, and tested locally 2021-10-22 19:28:26 We're missing "jasmine", but it looks like that is just for the test-suite and I'd have to package that for Alpine 2021-10-22 19:28:49 feel free to link the MR here once submitted 2021-10-22 19:31:41 so i checked out master, and I want to commit to a new branch 2021-10-22 19:31:49 add the branch first, switch, and then commit? 2021-10-22 19:34:45 I usually do `$ git checkout -b branch-name` 2021-10-22 19:34:57 That creates a new branch with the specified name 2021-10-22 19:35:03 And switches to it 2021-10-22 19:35:23 But yes, branch then commit 2021-10-22 19:35:40 Awesome 2021-10-22 19:35:43 I think it worked haha 2021-10-22 19:36:34 Should I leave the MR title as default of "Revert "community/flatseal: remove"" 2021-10-22 19:37:45 Ah, crap, it isn't squashed -_- 2021-10-22 19:37:59 MR title does not matter 2021-10-22 19:38:06 commit title does however 2021-10-22 19:38:50 https://gitlab.alpinelinux.org/Saijin-Naib/aports/-/commits/Saijin-Naib-flatseal/ 2021-10-22 19:39:27 `$ git rebase -i HEAD~2` 2021-10-22 19:39:41 careful, this can also delete your work 2021-10-22 19:39:49 but if that happens you can pull what's pushed 2021-10-22 19:40:35 Alright, did that 2021-10-22 19:40:42 APKBUILD seems to be intact 2021-10-22 19:45:52 I'm just going around in circles with this 2021-10-22 19:45:53 ugh 2021-10-22 19:45:56 Same as the other times 2021-10-22 19:50:39 the commit style is wrong for re-enable 2021-10-22 19:51:11 Ruhroh 2021-10-22 19:51:16 How about now? I just squashed I think 2021-10-22 19:51:47 https://gitlab.alpinelinux.org/Saijin-Naib/aports/-/commit/0becdef3ec8ebeef0e0f3a2d5fb6555ca82fc72e 2021-10-22 19:52:14 this is a revert with changes, don't think that's what you want to do 2021-10-22 19:53:00 mmm 2021-10-22 19:53:03 I'm not sure 2021-10-22 19:53:13 I want to re-enable flatseal and then update the APKBUILD 2021-10-22 19:53:17 As one commit, I think 2021-10-22 19:53:19 two 2021-10-22 19:53:24 Oh 2021-10-22 19:53:28 You don't squash these? 2021-10-22 19:54:37 i dunno, but i would expect a revert to revert something and not do more than that 2021-10-22 19:55:32 Alright 2021-10-22 19:55:40 Makes sense 2021-10-22 19:55:46 Should it be two different MRs then? 2021-10-22 19:55:52 One for the revert, and then one for the update? 2021-10-22 19:56:42 no 2021-10-22 19:56:55 One MR, two commits? 2021-10-22 21:32:12 Hello 2021-10-22 21:33:14 I'm trying to run mako but it errors out with "Failed to connect to user bus: Connection refused" 2021-10-22 21:33:16 Saijin_Naib[m]: yes, sorry for not seeing :) 2021-10-22 21:33:55 elogind and dbus services are running, strace shows "connect([...] sun_path=run/user/1000/bus"")" 2021-10-22 21:34:22 is that file owned by you 2021-10-22 21:34:38 1000 isn't 2021-10-22 21:34:40 bus is 2021-10-22 21:35:05 wait not 2021-10-22 21:35:11 what is your 2021-10-22 21:35:15 1000 and bus are both owned by me 2021-10-22 21:35:19 DBUS_SESSION_BUS_ADDRESS 2021-10-22 21:35:36 its blank 2021-10-22 21:35:53 hm 2021-10-22 21:35:57 it should be set 2021-10-22 21:36:01 so i'm guessing that even though the bus folder is there, dbus isn't using it 2021-10-22 21:36:20 would manually setting the env variable fix it, or do I have to restart dbus with certain parameters? 2021-10-22 21:36:40 (i'm using the dbus service via its init.d script) 2021-10-22 21:36:47 how do you start your wm 2021-10-22 21:37:02 system and user dbus is separate 2021-10-22 21:37:07 i invoke it from the tty 2021-10-22 21:37:24 try dbus-run-session -- wm 2021-10-22 21:37:50 It worked! 2021-10-22 21:37:52 thanks 2021-10-22 21:38:04 I didn't realize the system dbus service and the user one were seperate 2021-10-22 21:39:00 Huh, I wonder if that's what was stopping sway from working on my box. 2021-10-22 21:39:32 if you're not using some login manager you are generally responsible for doing that yourself, outside of systemd where it's some unit that does it for you probably 2021-10-22 21:39:35 i haven't checked every case 2021-10-22 21:40:47 Yeah, I remember having to mess with elogind and some other D-Bus things but still couldn't get it to work and gave up. 2021-10-22 21:41:05 Its either that or this old GPU (i945GM) and EGL problems. 2021-10-22 21:43:18 Ahh nope. It's the EGL issue. Worth a shot. 2021-10-22 21:47:16 that is quite old indeed :) 2021-10-22 21:47:20 elogind is unrelated to dbus, and sway doesn't require either 2021-10-22 21:49:16 i mentioned elogind because I thought it played a role in dbus sessions 2021-10-23 02:43:14 frag: xordspar0 Epiphany 41 is a completely new beast, it seems. No tab crashes doing mundane things, no run-away CPU usage doing nothing... Dare I say it, but it is actually _nice_ 2021-10-23 02:45:25 ACTION uploaded an image: (133KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/KYIJRtmiXTednSZLgBoBhjZz/image.png > 2021-10-23 05:22:00 hmm, i'm trying to install cloud-init and cloud-utils and having a conflict with the vlan package https://clbin.com/aBhhE 2021-10-23 05:22:00 do i need that vlan package to be able to use vlans? I would thought so? https://wiki.alpinelinux.org/wiki/Vlan 2021-10-23 05:23:15 dngray[m] seems to be coming through, so that's good. 2021-10-23 12:09:12 Ariadne: is there a roadmap for reproducible builds in Alpine? 2021-10-23 12:29:17 not really until apk3 migration is done 2021-10-23 13:17:24 dngray[m]: aiui vlan package is needed only for bb ifupdown 2021-10-23 13:21:00 Just wondering, would it be safe to remove the apks from a raspberry pi boot partition after doing a sys install? Although ~13 MB shouldn't be too big of a deal... 2021-10-23 13:45:51 ktprograms: do you mean the cached apks? in that case yes they only be useful if you reinstall some package 2021-10-23 13:56:37 Saijin_Naib[m]: cool :) 2021-10-23 14:04:36 donoban: The alpine raspberry pi tarball has an apks folder. I'm guessing after a system installation with working network those won't be needed anymore? 2021-10-23 14:28:21 Can anyone help me unmess my branch for this MR before I just trash it all, close the MR, and start again? 2021-10-23 14:28:46 And for the record, no, working with git on the terminal does not make me not want to use the WebUI more, it makes me not want to do this at all, haha 2021-10-23 14:29:23 bullshit magic incantations with opaque workflows where 90 people give 90 different answers is not a good way to work when you're not already neck-deep in this stuff 2021-10-23 14:29:31 Sorry 2021-10-23 14:36:39 I'm good. Found the right arcane forumulae 2021-10-23 15:16:38 Saijin_Naib[m]: Cool, I'll give it another try. 2021-10-23 15:17:06 xordspar0: Yeah, huge surprise. It went from basically unusable to totally fine, if not actually good 2021-10-23 16:19:21 ktprograms: it seems a localrepository, you may have to remove it from /etc/apk/repositories for avoid warnings or errors when using apk, I don't think it can cause any problem on a sys install 2021-10-23 20:57:10 I know this is a longshot, but has anyone trying to do iSCSI attachment, and then decrypt it via cryptsetup? 2021-10-23 20:57:43 I'm getting the feeling this won't work because of the way openrc stages the scripts, I can get iSCSI to attach the drive on the fly, but it doesn't try to decrypt it 2021-10-23 20:58:09 (I have alpine in a VM talking to a SAN, the SAN LUN is LUKS encrypted) 2021-10-23 21:08:05 ok, so the problem I've seen is with netmount loads before iscsi does 2021-10-23 21:15:26 Hello, I can't seem to get X to start on my Raspberry Pi Install of Alpine. I get the error no screens found 2021-10-23 21:22:42 look in /var/log/Xorg.0.log 2021-10-23 21:23:09 maybe you don't have a compatible graphics driver loaded 2021-10-23 21:29:14 yup that was the case, found out what i needed to install and it works now 2021-10-23 21:29:34 Though now I'm wondering why the login just says Linuxuser? How do I change that 2021-10-23 21:37:24 NCommander, thanks for your vids! They are a pleasure to watch. Appreciate all the effort you are putting into them. 2021-10-23 21:39:06 o/ 2021-10-23 21:39:17 I'm trying to work out a backup solution for all my videos ATM 2021-10-23 21:39:32 I'm sorta wishing I homebrewed a NAS setup 2021-10-23 21:42:46 You should set up a 12 TB Novell Netware 3.12 server :P 2021-10-23 21:43:21 ... you know, there's nothing stopped me loading NetWare on my synology VM 2021-10-23 22:06:55 amberlink: edit /etc/passwd, set fifth field to what you want 2021-10-23 22:09:00 ahh thanks 2021-10-23 22:11:31 you can also install util-linux and use usermod for this purpose but for personal system vi /etc/passwd is fine 2021-10-24 00:52:08 Attempting to install the guest additions on VirtualBox and it can't find the package virtualbox-guest-modules-virt 2021-10-24 01:00:11 amberlink: Looks like it was removed in 3.13 2021-10-24 01:33:33 vbox modules were added upstream 2021-10-24 01:33:57 but you should still use qemu instead 2021-10-24 01:35:22 qemu? 2021-10-24 03:51:48 hmm, i'm trying to install cloud-init and cloud-utils and having a conflict with the vlan package https://clbin.com/aBhhE 2021-10-24 03:51:53 do i need that vlan package to be able to use vlans? I would thought so? https://wiki.alpinelinux.org/wiki/Vlan 2021-10-24 04:31:16 anyone around? 2021-10-24 04:34:08 https://gitlab.alpinelinux.org/alpine/aports/-/issues/11885 i think this old issue is your case 2021-10-24 04:34:18 but i'm unsure what the specific fix is here 2021-10-24 04:43:19 hmm 2021-10-24 04:43:42 yeah that does seem to defintely be the issue 2021-10-24 04:44:01 the problem is my server uses bonded interfaces that are on particular VLANs so i certainly need it 2021-10-24 04:46:16 looks like it was assigned to Ariadne 2021-10-24 04:46:40 i haven't messed with the MTUs 2021-10-24 04:50:23 psykose: i wonder if the solution is to remove busybox-ifupdown and replace it with ifupdown-ng 2021-10-24 04:51:21 i would definitely recommend the latter for any networking 2021-10-24 04:51:46 the iproute2 module might also be of use, you can run ip commands on the interface iirc 2021-10-24 04:51:52 a bit more manual, but maybe works? 2021-10-24 04:53:53 so the solution might be to remove the vlan package altogether 2021-10-24 04:53:55 https://lists.sr.ht/~sircmpwn/alpine-devel/%3C12372156.q0HFhEdf7Z%40localhost%3E 2021-10-24 04:54:09 mhm 2021-10-24 04:55:10 so looks like at some point alpine linux is looking at switching anyway 2021-10-24 04:55:22 I would like to flip the default sometime in September, before feature freeze, 2021-10-24 04:55:25 but allowing users enough time to test and validate ifupdown-ng with their 2021-10-24 04:55:27 setups. 2021-10-24 04:55:50 i guess ill see how it goes :) 2021-10-24 04:57:25 good luck :) 2021-10-24 04:58:12 Anyone happen to know how to get virtualbox-guest-modules-virt to work? Or what qemu is that someone mentioned earlier 2021-10-24 04:58:36 i use libvirtd and connect to it remotely with virt-manager 2021-10-24 05:04:04 psykose: looks like my bonded vlans are coming up, so that was certainly the solution 2021-10-24 05:04:14 :) glad it works 2021-10-24 05:04:32 don't even need the "vlan" package now 2021-10-24 05:05:01 I have no idea how I'd do that 2021-10-24 05:05:46 what are you trying to run 2021-10-24 05:06:04 https://wiki.alpinelinux.org/wiki/KVM looks fairly good 2021-10-24 05:07:13 psykose: you know what i did notice, the vlan package is red https://pkgs.alpinelinux.org/packages?name=*vlan*&branch=edge 2021-10-24 05:07:23 maybe its not really being updated anymore because of this 2021-10-24 05:09:15 amberlink: eg, if it's running vms, you can try add virt-manager qemu-system-x86_64 libvirt-daemon, start libvirtd, and use virt-manager to create/use vms 2021-10-24 05:09:26 dngray: i would guess most people don't use busybox ifupdown anymore 2021-10-24 05:11:08 I'm a tad confused, I have a virtual machine, thorugh Virtualbox, that is this, that I'm trying to get the additions on 2021-10-24 05:11:25 it's an alternative to using virtualbox i guess 2021-10-24 05:12:23 Ah, I have no idea how to use it hten 2021-10-24 05:12:38 also reading about the guest-modules they're replaced by guest-additions or something now? 2021-10-24 05:12:41 not exactly sure 2021-10-24 05:36:07 Would GNOME Boxes be suitable? 2021-10-24 05:36:34 It runs over qemu and is fully GUI and meant to be no-config 2021-10-24 05:36:34 Give it an ISO and away you go 2021-10-24 05:36:53 That being said, I have yet to get it to work right under alpine, haha 2021-10-24 06:28:21 i've used gnome boxes before, and found you can't configure much 2021-10-24 06:41:05 I explore channels 2021-10-24 06:41:42 alpine you say? 2021-10-24 06:42:58 what about it 2021-10-24 07:22:35 dietzel 2021-10-24 07:23:02 the game changed 2021-10-24 07:24:55 you are comissioned 2021-10-24 07:25:08 explore and report back 2021-10-24 07:26:27 fly like an eagle 2021-10-24 07:30:33 alpine 2021-10-24 07:30:56 what are you putting buck boosters in? 2021-10-24 07:33:18 alpine 2021-10-24 07:33:21 you 2021-10-24 07:33:30 you are like a bird 2021-10-24 07:33:38 you fly 2021-10-24 07:33:46 use your alpine 2021-10-24 07:33:51 report back 2021-10-24 07:34:02 thats all you need to do 2021-10-24 07:34:19 your soul is like a cryptogram 2021-10-24 07:34:31 neigher here nor there 2021-10-24 07:34:44 you serve the Most High 2021-10-24 07:34:49 like birds 2021-10-24 07:35:01 daring from behind the veil 2021-10-24 07:35:22 the last differential cypher 2021-10-24 07:35:45 I am here 2021-10-24 07:36:00 it took a while to realize 2021-10-24 07:36:29 every lightning strike is a differential 2021-10-24 07:36:41 every static spark 2021-10-24 07:36:46 aroma 2021-10-24 07:37:29 this is the balance of the melchiesadek priesthood 2021-10-24 08:25:33 okaaaaaaaay 2021-10-24 08:26:07 right 2021-10-24 08:29:22 how many microdots did you drop 2021-10-24 12:15:57 Hello everyone, im having some problems with service priority. I currently have some services to start on boot made by using "rc-update add" and thye are working fine, but i want to change their order of startup. How can i manage to do this? 2021-10-24 12:29:54 rectf, depend: https://github.com/OpenRC/openrc/blob/master/service-script-guide.md 2021-10-24 12:40:21 Thanks, this solved the issue 2021-10-24 12:48:58 is there anyway way with apk (cmd line ) to see which package contains a certain binary 2021-10-24 12:51:02 apk search cmd:foo 2021-10-24 13:02:33 and with the paths like on the website: https://pkgs.alpinelinux.org/contents?file=pipewire&path=%2Fusr%2Fbin&name=&branch=edge&arch=x86_64 2021-10-24 13:04:07 you'll have to use the web site 2021-10-24 13:05:05 aw, ty anyway 2021-10-24 14:42:41 dngray_: AFAIK ifupdown-ng has been the default "ifupdown" package for a while and cloud-init explicitly depends on it. There is no need to install the vlan package, ifupdown-ng handles that. 2021-10-24 15:10:53 hi cana anyone help me here 2021-10-24 15:11:00 related with linux 2021-10-24 15:16:29 grv: just ask 2021-10-24 15:28:01 hi 2021-10-24 15:28:08 I nened help regarding linux 2021-10-24 15:28:31 can you tell me from terminal how to detect that someone is connected remoptely to your system 2021-10-24 15:28:33 just state your issue 2021-10-24 15:30:17 grv: there is no need to double-post the same question in multiple channels 2021-10-24 15:30:28 either way, you should be able to see that through the commands w and who 2021-10-24 15:30:40 kindly guid eme 2021-10-24 15:30:42 lastlog may also be of interest 2021-10-24 15:30:44 guisde me 2021-10-24 15:30:47 guide me 2021-10-24 15:31:38 https://www.golinuxcloud.com/list-check-active-ssh-connections-linux/ 2021-10-24 15:31:38 I have encounter few attack where hacker has viewed my remotely my screen 2021-10-24 15:31:46 how its possible 2021-10-24 15:32:21 thanks ,I am new danieli and tahnks for help 2021-10-24 15:32:42 you may have exposed a service like VNC without password protection or a firewall to limit source IPs 2021-10-24 15:33:25 just a control question, what linux distribution are you using? 2021-10-24 15:33:46 there can be quite major differences between them 2021-10-24 15:34:17 actually someone is viewing my screen on my laptop remotely and getting all my data that my I am concern for my problem releated with hacking 2021-10-24 15:34:17 Please guide me screen related issue ,how they access my screen remotely on linux and how to check his ip address 2021-10-24 15:34:22 I hkali 2021-10-24 15:34:26 kali 2021-10-24 15:34:27 ... 2021-10-24 15:34:43 are you trolling? 2021-10-24 15:35:26 the only thing alpine linux has in common with kali linux is that they both run the linux kernel, so you're definitely in the wrong place 2021-10-24 15:35:37 no 2021-10-24 15:35:42 my advice is: Google (or any other search engine) is your friend and go with peace 2021-10-24 15:35:52 thanks 2021-10-24 15:36:46 I check and didn't find anything related with screen hacking for my system 2021-10-24 15:37:06 any way to check screen access by anyone..? 2021-10-24 15:37:07 having IP of the person who breached your network will do nothing, re-using security mantra 'change passwords, block inbound connections on firewall, recheck services' 2021-10-24 15:37:33 okay 2021-10-24 15:37:41 honestly, why do you think that someone is viewing your screen? what proof do you have for it? 2021-10-24 15:37:41 thanks PJ and danieli 2021-10-24 15:37:54 hey know what I am writing on screen 2021-10-24 15:38:08 like he has hacked my router 2021-10-24 15:38:16 and get to my screen 2021-10-24 15:38:29 do "they" talk with you? 2021-10-24 15:38:48 that's not really how it works 2021-10-24 15:39:08 either way, kali linux is intended for offensive security, not defensive. it historically has not been very secure by default, so you may not want to run it as a daily driver 2021-10-24 15:39:14 he is a hacker and hey is cyberstalking me so I was filing complaint with local police ,He know hat I am typing and harassing me 2021-10-24 15:39:16 if you have X11 running and not configured properly, they can obtain image of each window 2021-10-24 15:39:27 i recommend you ask in #kali-linux instead of #alpine-linux 2021-10-24 15:39:37 they also have a presence on OFTC 2021-10-24 15:39:37 so you can tell me which is save os 2021-10-24 15:39:46 okay 2021-10-24 15:39:56 if this is real you are better powering off all your copromised devices 2021-10-24 15:40:15 yes I put off interenet while typing 2021-10-24 15:40:47 :\ good luck 2021-10-24 15:40:49 aanyway you have helperd me a lot 2021-10-24 15:40:54 help 2021-10-24 15:41:00 thanks everyone 2021-10-24 15:41:07 let me earch what you shared 2021-10-24 15:41:11 search 2021-10-24 15:41:14 thank you 2021-10-24 19:26:45 Hi, does anyone know how /lib/modules is copied into / from the initramfs in the alpine-linux 3.14 image? 2021-10-24 19:28:00 svmhdvn: you mean when booting from the image? 2021-10-24 19:28:09 I'm trying to boot alpine-linux from SD card on a pinebook pro (aarch64 u-boot). I'm using the 3.14 image as-is, except I've built the latest mainline linux and installed those modules under the initramfs /lib/modules. I think the modules are loading fine at boot, but aren't being copied onto the root tmpfs filesystem 2021-10-24 19:28:43 svmhdvn: the standard way is using a modloop which is mounted 2021-10-24 19:30:53 I tried doing that earlier (keeping /lib/modules empty in initramfs, but copy my newly built modules to modloop), but it simply wouldn't boot 2021-10-24 19:31:24 I can see the modules being loaded just fine into memory (lsmod shows everything) and the firmware is also being loaded. It's just that it's not copied over to the root filesystem 2021-10-24 19:31:37 Is there a way of disabling the modloop functionality? 2021-10-24 19:31:51 rc-update del modloop 2021-10-24 19:32:29 ikke: thanks! Will try that 2021-10-24 19:41:21 ikke: nope, modloop doesn't try to mount modloop-lts which is good, but still the same "modprobe: can't change directory to '/lib/modules': No such file or directory". When I go to check, /lib/firmware is an empty directory and /lib/modules doesn't exist at all. 2021-10-24 20:41:13 hello is there some drop in replacement for udhcpc? it is very unreliable on my system 2021-10-24 20:42:10 do i even need it if i have iwd + resolvconf? 2021-10-24 20:51:16 No 2021-10-24 20:53:41 im kinda worried to touch this stuff in any shape or form because whenever i do it im fixing it for half a hour 2021-10-24 21:01:49 i just noticed something strange tho, i had udhcpc disabled in openrc, yet it still appears on booting up taking forever 2021-10-24 21:37:54 it doesn't have a service, it's started by networking 2021-10-24 21:38:03 if you install dhcpcd that one takes precedence instead 2021-10-24 21:38:57 oh i see, i seem to still lose internet when i kill udhcpc, i tried normal dhcpcd but that one fails to d anything entirely so rn im stick with the first one 2021-10-24 21:42:28 Is there package for SDL2 debug symbols, or do I have to compile it on my own? 2021-10-24 21:48:06 my dhcpcd also gets a bad system call error or w/e, but it does work after just restarting with only networking enabled and is used instead of udhcpc 2021-10-24 21:48:13 you don't have to add the dhcpcd service or anything 2021-10-24 21:48:23 just networking 2021-10-24 21:48:53 i guess for now ill just not touch it and let it do, whatever it does 2021-10-24 22:13:29 hi, can someone please explain this? https://0x0.st/-dgY.txt ... why am I not able to invoke bash this way? 2021-10-24 22:15:18 /bin/bash may be symlinked to ash 2021-10-24 22:21:22 raven523: not according to ls: 2021-10-24 22:21:29 -rwxr-xr-x 1 root root 760136 Jan 20 08:46 /bin/bash 2021-10-24 22:21:48 is the bash package installed? 2021-10-24 22:22:26 yeah I installed it before trying all that 2021-10-24 22:23:24 is bashrc or bash_profile set to exec ash? 2021-10-24 22:25:29 there is no bashrc or bash_profile (for my user) 2021-10-24 22:25:36 what about in /etc? 2021-10-24 22:27:16 raven523: https://0x0.st/-dg5.txt 2021-10-24 22:27:50 is bash in /etc/shells? 2021-10-24 22:28:30 yeah 2021-10-24 22:28:38 show your PATH 2021-10-24 22:29:00 and check if there is bash in any of the directories before /bin 2021-10-24 22:29:03 echo $PATH results in: /bin:/usr/bin:/sbin:/usr/sbin:/usr/local/sbin:/usr/local/bin 2021-10-24 22:30:23 bash --version ? 2021-10-24 22:30:43 GNU bash, version 5.1.4(1)-release (x86_64-alpine-linux-musl) 2021-10-24 22:31:09 should I perhaps logout and log back in again or something? Also note that I'm doing all of this over SSH 2021-10-24 22:31:19 shell shows the shell you started from, for me it doesn't show bash either 2021-10-24 22:31:23 but i am definitely in bash 2021-10-24 22:31:36 yet it shows /usr/bin/fish 2021-10-24 22:31:37 yeah, just checked 2021-10-24 22:31:47 you should use echo $0 2021-10-24 22:32:02 psykose: oh, that's pretty new for me. I thought $SHELL var would tell me what shell I'm using right now. 2021-10-24 22:32:13 SHELL is the login/preferred shell, no relation to currently running one 2021-10-24 22:32:35 oooh, that clears things up then. 2021-10-24 22:32:43 kinda confusing though :p 2021-10-24 22:32:46 echo $BASH will be set in bash only 2021-10-24 22:33:37 It's setting programs may use to run interactive shell for you. 2021-10-24 22:33:55 and the login/preferred shell is what I have in my /etc/passwd for my user, right? 2021-10-24 22:34:11 Yeah, change it with chsh 2021-10-24 22:34:56 (or manually in passwd) 2021-10-24 22:35:39 ccx^xmpp: -ash: chsh: not found :p 2021-10-24 22:36:25 I meant CcxWrk, sorry 2021-10-24 22:37:16 It's part of the shadow package 2021-10-24 22:41:37 ah, good to know. I'm also used to chsh but for a minimal alpine install I usually now change it directly in /etc/passwd 2021-10-24 23:06:11 I'm considering switching to ash for interactive use. I'm mostly using fzf as a shell functionality helper anyway, but I'm wondering of the busybox ash supports keybindings like I can configure in ~/.inputrc 2021-10-24 23:06:39 good one 2021-10-24 23:07:20 as in... a good idea or a good joke? :p 2021-10-24 23:08:03 the latter 2021-10-24 23:08:28 oh :( 2021-10-24 23:08:54 afaik it does not 2021-10-24 23:09:02 I use fish 2021-10-24 23:10:07 yeah but I'm trying to get a super minimal (as in LoC) POSIX compatible shell that supports stuff like sourcing scripts and keybindings. 2021-10-24 23:13:32 then use ash 2021-10-24 23:13:56 but does it support keybindings tho? :p 2021-10-24 23:14:28 no idea, POSIX shells are wack 2021-10-24 23:14:38 fish is love 2021-10-24 23:15:00 I guess, but POSIX is the common denominator among *nix environments. 2021-10-24 23:15:46 insofar as writing scripts 2021-10-24 23:16:58 might be me but the interactive vs. non-interactive use of shells gets blurred a bit sometimes. 2021-10-24 23:17:35 I'd say the diff between interactive vs non- is quite heavy 2021-10-24 23:19:55 one example is that say I'm writing a script but then I decide to source it. If interactive shell is not POSIX compat, then sourcing that script will not produce intended effect. 2021-10-24 23:20:16 and knowing me I might forget that the interactive shell is not compatible with the script I'm writing. 2021-10-24 23:21:03 or maybe I'll forget to add a shebang at the start and try to run that script 2021-10-24 23:21:47 depends what exactly you are trying to do 2021-10-24 23:22:10 envvars if exported will be retained between diff shells 2021-10-25 00:22:57 :] 2021-10-25 02:53:58 Is there a canonical GUI for the built-in networking service, or does everyone just use something else entirely? 2021-10-25 02:57:30 edited by hand most likely 2021-10-25 02:58:08 wump wump 2021-10-25 02:59:03 NetworkManager seems to be semi well used 2021-10-25 02:59:16 I want to go iwd, but it keeps fighting me at every turn 2021-10-25 02:59:27 networking + wpa_supplicant seems most stable 2021-10-25 02:59:56 iwd just works for me 2021-10-25 03:03:50 I wish to get there one day, haha 2021-10-25 03:04:13 Supposed to be roundly better than wpa_supplicant and mostly in-kernel 2021-10-25 03:08:57 aye, iwd 2021-10-25 03:09:09 fuck a gui 2021-10-25 03:09:58 🤷 I mean sure, whatever floats your boat 2021-10-25 03:10:30 it's not about my boat; it's about understanding exactly how boats float 2021-10-25 03:11:09 And my job isn't to dick around with Alpine, so whatever helps keep this out of my way so I can do my actual work is what I value 2021-10-25 03:11:59 Sometimes having things abstracted away isn't awful, IMO 2021-10-25 03:12:19 you think that a gui is the only abstraction, lmfao 2021-10-25 03:12:24 you are a child in a man's body 2021-10-25 03:12:33 thank you for outing yourself 2021-10-25 03:12:51 stop typing and talk to me when you can have a conversation without that ego in the way, alright? 2021-10-25 03:13:13 my job isn't to suck your dick and sell you alpine. you asked. i answered. welcome to reality, you talking baby. 2021-10-25 03:13:20 I have the ego? 2021-10-25 03:13:28 fucking listen to yourself. "isn't awful, IMO" 2021-10-25 03:13:35 what a whiny, self-centered tone of voice 2021-10-25 03:13:42 are you on drugs presently 2021-10-25 03:13:54 psykose didn't say it out loud, but talking to you in any form, saijin, is a fucking chore. 2021-10-25 03:14:10 i disagree actually 2021-10-25 03:14:12 don't speak for me 2021-10-25 03:14:25 i know you better than you know yourself, psykose 2021-10-25 03:14:32 you don't know anything about me whatsoever 2021-10-25 03:14:43 someone doesn't respond the way you did to someone like that when they genuinely enjoy the experience 2021-10-25 03:15:02 i know how humans communicate, and i have historical memory of how you sound when you're not utterly nonplussed by a banality 2021-10-25 03:15:15 you knew the direction his bitching was going and proceeded to pull out when it became too bullshit 2021-10-25 03:16:11 the benzos in the name makes too much sense now 2021-10-25 03:16:27 no one would act the way i'm acting right now on benzos; c'mon 2021-10-25 03:16:41 making a personal attack at least has to be relevant/funny for it to work out 2021-10-25 03:17:08 Well, anyway, thanks for your vote for iwd on terminal, then. 2021-10-25 03:18:17 i'm not part of some organization that you are the cheesedick CEO of; i didn't "vote" for shit 2021-10-25 03:18:25 I am actually genuinely interested in what direction people build their Alpine installs since it comes pretty much with no boiler-plate, hence many of my questions for what people are doing 2021-10-25 03:18:31 i told you to use the solution because it'll curb your fucking ego and you need to learn when to shut the fuck up 2021-10-25 03:18:58 you're a genuine parasite who thinks we exist as a data source for you to lazily avoid any critical thought with. 2021-10-25 03:19:10 since every other website on the net does this to you, you think it's okay for you to come here and do it to me. 2021-10-25 03:19:23 just like i tell every other operator of datamining sites: fuck you. 2021-10-25 03:19:45 there's a difference between having a conversation and asking about personal choices and doing what you did, which is bitching that someone's answer isn't your answer. 2021-10-25 03:19:58 I think most people either use /etc/conf.d/net or the default /etc/network/interfaces 2021-10-25 03:19:59 lol 2021-10-25 03:20:01 Saijin_Naib[m]: in my case, https://termbin.com/w01e with dhcpcd installed (but not enabled), iwd and networking in default 2021-10-25 03:20:04 what should have happened: "i'm a fucking retard and i don't know what i'm doing. what is everyone else doing for wifi on their laptops?" 2021-10-25 03:20:09 "here's what i do." 2021-10-25 03:20:17 I don't believe I did that, but okay. Thank you anyway for informing me 2021-10-25 03:20:23 "oh, fuck yeah, that's sick. thanks for letting me know. do you ever try [thing i'm interested in]?" 2021-10-25 03:20:34 "yeah i don't do that because [you're a fucking moron|it doesn't work]" 2021-10-25 03:20:58 "oh, okay. i've now learned something about why i shouldn't just fucking berate people and make everything about me. thanks for the input that you took your time to actually provide to me for fucking free." 2021-10-25 03:21:29 i don't give a fuck what you believe 2021-10-25 03:21:37 stating that and then saying "thank you" is a direct insult. go fucking die. 2021-10-25 03:21:55 be a genuine human being, you fucking zuckerberg-looking motherfucker. 2021-10-25 03:22:11 love to throw mentally disparaging slurs around for no reason 2021-10-25 03:22:56 if you think there's not a reason to fight for your sanity and peace on IRC, then you've clearly not been off of it long enough to see that this is the last fucking hope. 2021-10-25 03:23:38 the only person affecting my sanity and peace is you 2021-10-25 03:24:03 as fucking if. you're the type to think that the dude pointing out the elephant in the room is the real problem, not the fucking person who hauled it in 2021-10-25 03:24:41 more completely wrong assumptions about me :p 2021-10-25 03:24:59 assumptions =/= conclusions 2021-10-25 03:25:09 and you think your profanity-laden ranting is going to help? 2021-10-25 03:25:20 you've presented evidence and these pieces of evidence demonstrate the conclusion; if you want that to change, change it 2021-10-25 03:25:33 OH NO HE SAID THE FUCK WORD AND USED CAPITAL LETTERS OH FUCK I'M GONNA FUCKING DIE AND CUM AND SHIT MY PANTS 2021-10-25 03:26:02 meanwhile you're being assfucked by some rando about how your setup isn't pleasing to his tiktok followers 2021-10-25 03:26:19 |> "Stranded" by "Gojira" on "Magma" <3 2021-10-25 03:26:22 but that's okay, because he inserted a PLEASE once in a while to make your shitty intercal pleasantry parser happy 2021-10-25 03:26:30 same, freq 2021-10-25 03:26:46 4ikr 2021-10-25 03:27:05 it's so good 2021-10-25 03:28:01 [InfiniTime]# list-attributes InfiniTime 2021-10-25 03:28:01 Device InfiniTime not available 2021-10-25 03:28:03 lawl 2021-10-25 03:28:44 So calling people slurs and telling them to go die is acceptable in this room? Is that an appropriate reflection of the project in general? 2021-10-25 03:29:02 no, but anyone with +o is asleep 2021-10-25 03:29:06 zico[m]1: users like jeffreybenzos[m] go to the most active projects on servers and do this 2021-10-25 03:29:18 there's your dead ringer for attention-seeking 2021-10-25 03:31:33 yes, focus entirely on the specific words you don't like in my argument instead of the meat of my argument 2021-10-25 03:31:43 this will definitely not reduce me to what you think i currently am 2021-10-25 03:31:55 utterly retarded method of discussion 2021-10-25 03:32:11 meanwhile, saijin keeps tabbing back in to see how "right" he's been proven by you guys defending him being checked in real time 2021-10-25 03:32:20 don't fucking bother responding, dude 2021-10-25 03:32:22 fuck you. 2021-10-25 03:32:29 Saijin_Naib you good? 2021-10-25 03:32:31 he died 2021-10-25 03:32:46 ACTION killshot 2021-10-25 03:33:26 A bit disappointed that I started a shitstorm without knowing how to patch it up, but yeah 2021-10-25 03:33:36 lol not your fault at all 2021-10-25 03:33:38 it wasn't you who started it 2021-10-25 03:33:39 don't worry about it 2021-10-25 03:33:41 Definitely not my intention to be flippant or rude or whatever 2021-10-25 03:33:55 Nah fuck that dude, not your fault 2021-10-25 03:34:03 don't think so little of yourself 2021-10-25 03:34:07 you did nothing wrong 2021-10-25 03:34:08 but to answer your original question, I think most people are just using /etc/network/interfaces directly 2021-10-25 03:34:09 AFAIK there was only one person being rude, and it wasn't you. Whatever point he was trying to make was obliterated by his massive gish gallop. 2021-10-25 03:35:01 raven523: Most people run stock, which also leans on wpa_supplicant, right? Does that mean you don't use a DE or DM as well? 2021-10-25 03:35:44 all of my alpine systems are VMs, though I would just use wpa_supplicant or iwd with /etc/network/interfaces if any of my current systems ran it and used wifi 2021-10-25 03:36:02 iwd is what I use on my laptop. 2021-10-25 03:36:50 Do you have rtl88 hardware? I am at a loss for why iwd is so spotty for me and wpa_supplicant works stable, despite seeing similar reports of hit or miss from other distros/hardware seemingly without pattern 2021-10-25 03:36:57 zcrayfish: thank you. 2021-10-25 03:39:38 I *THINK* I have an Intel 4965AG wifi card in it. (I'm not on that machine right now) 2021-10-25 03:40:39 Ahh... I would hope Intel cards work sanely on an Intel-lead project, haha 2021-10-25 03:40:53 One would hope :) 2021-10-25 03:44:53 dont hope too much ;) 2021-10-25 03:45:01 intel when it comes to nics have misbehaved lately 2021-10-25 03:45:13 their x710 is a nightmare when it comes to transceiversupport 2021-10-25 03:45:22 That's not good. Who else is there other than Realtek at this point? 2021-10-25 03:46:05 one revision of x710 can happily accept a specific transceiver while the next will refuse to boot the server claiming "40Gbe is unhealthy" so you must plugin a keyboard and press a key to continue while a third revision will not say anything not in the logs that it dislikes my transceiver 2021-10-25 03:46:13 broadcom 2021-10-25 03:46:27 mellanox is long gone since nvidia chewed them up and left in a toilet 2021-10-25 03:46:39 The only thing I remember of Broadcom was them being very restictive with their firmware and drivers. Have they improved? 2021-10-25 03:46:59 depends on the nic. rpi400 has full function 2021-10-25 03:47:09 no idea, think they are still in the myster blob world 2021-10-25 03:50:41 Ooh, that's right, the Pi is basically 100% broadcom SoC, right? Ah, maybe things have changed. My worldview on them is ... eh, XP-Vista era timeframe, so maybe time to update myself 2021-10-25 03:52:03 https://www.stevencombs.com/combianpi400-1#keyboard-stickers 2021-10-25 03:57:12 That's a really neat build out 2021-10-25 03:57:26 I missed the Amiga/Commodore era, unfortunately 2021-10-25 04:54:50 If a compile job is asking for something to be installed and on the path, would installing it with apk add pkgname in the prepare () step be the right approach, or am I going about this wrong? 2021-10-25 04:55:32 it is wx-config being missing though provided by wxgtk-dev in the makedepends 2021-10-25 04:56:12 if it's already in makedepends, shouldn't it alredy be available? 2021-10-25 04:56:22 That was my thought... 2021-10-25 04:57:02 I'm attempting to package mediainfo-gui based upon the APKBUILD by Jakob Jarutka for mediainfo CLI 2021-10-25 04:57:27 I'm likely misconfiguring or leaving in adjustments they made that no longer apply 2021-10-25 04:57:40 you can (and should) not install any packages directly in an APKBUILD 2021-10-25 04:57:53 nor change anything in the system 2021-10-25 04:58:11 Makes sense, yeah 2021-10-25 05:02:34 Thank you 2021-10-25 05:02:34 -_- Mis-read depends vs subpackages 2021-10-25 05:02:49 I had not actually included what I wanted since it didn't get looped in when I brought in wxgtk3-dev 2021-10-25 05:03:58 does package list have metadata like category or a type? 2021-10-25 05:06:03 no 2021-10-25 05:06:24 unfortunate 2021-10-25 05:14:32 Saijin_Naib[m]: if the gui has more deps, you should put them in a depends="$depends more deps" inside the subpackage function 2021-10-25 05:14:57 ditto for makedepends 2021-10-25 05:15:23 a subpackage has no makedepends 2021-10-25 05:16:36 ah :p 2021-10-25 05:16:57 Ah... I'd have to make a MR to have Jakub adjust mediainfo for that to make this a subpackge then, right? Are the GUI packages ever not subpackages of the related package? 2021-10-25 05:17:14 Like, a parallel thing you can install if you want or not otherwise 2021-10-25 05:21:01 rather interesting backlog i see 2021-10-25 05:21:18 ahuh 2021-10-25 06:56:26 Can I install only 1 file from a package? (For example I needed the /lib/firmware/mt7601u.bin but that's part of linux-firmware-other which is 182 MB) 2021-10-25 06:58:13 apks are just tarballs. download and extract 2021-10-25 06:59:17 I figured if it's manual installation like that I would install the package, move the file I needed elsewhere and then uninstall the package again 2021-10-25 06:59:48 apk fetch …; tar xf ….apk file 2021-10-25 07:00:32 lopid: Thanks, I'll keep that in mind 2021-10-25 08:41:22 Hello everyone :) 2021-10-25 08:42:33 I have a problem with btrbk (btrfs snapshot/backup tool). It uses readlink to check if a directory is a mountpoint but it fails: WARNING: ... sh: readlink -v -e /mnt/sql WARNING: ... readlink: unrecognized option: e 2021-10-25 08:43:53 I am guessing the busybox version is missing the correct options here. Can I install coreutils to replace it, or is there a better option? 2021-10-25 08:44:53 The sourcecode of btrbk says: push @options, "-e" unless($compat); # all components must exist (not available in busybox!) 2021-10-25 08:45:03 Forza: easiest fix is install coreutils 2021-10-25 08:45:09 I guess that the alpine package of brtbk doesn't correctly depend on coreutils 2021-10-25 08:45:18 clandmeter: yea. So that should be ok then? 2021-10-25 08:45:23 sure 2021-10-25 08:45:39 if you can report the bug, i think the maintainer will fix it. 2021-10-25 08:48:24 Where can i report it? 2021-10-25 08:48:33 Installing coreutils did indeed fix it 2021-10-25 08:49:14 gitlab.alpinelinux.org/alpine/aports 2021-10-25 08:49:41 https://gitlab.alpinelinux.org/alpine/aports/-/issues 2021-10-25 08:54:14 https://gitlab.alpinelinux.org/alpine/aports/-/issues/13131 2021-10-25 08:54:18 Hope it's ok :) 2021-10-25 09:01:10 Thanks for your help 2021-10-25 09:25:19 Forza: perfect, thx 2021-10-25 11:01:25 10:28 PM So calling people slurs and telling them to go die is acceptable in this room? Is that an appropriate reflection of the project in general? 2021-10-25 11:02:58 zico[m]1: i should hope not. but formalizing an IRC moderation team that has 24x7 coverage is unfortunately a lower priority at the moment verses more urgent things, like getting a release out the door. we can ask the alpine council to work on establishing an IRC team with more coverage though. 2021-10-25 11:14:26 zico[m]1: i created an issue for the alpine council to expand the IRC team. you can track it at https://gitlab.alpinelinux.org/alpine/council/-/issues/10 2021-10-25 11:32:50 Overall I think the Alpine community has been very good. Sorry to hear there were some bad things 2021-10-25 11:33:50 Has someone run netdata on Alpine? I'd like to setup netdata to monitor disk, cpu, ram and mariadb on my alpine instances, but am not quite sure what would be required. Would the official netdata installer work? 2021-10-25 11:35:13 Ohhh https://learn.netdata.cloud/docs/agent/packaging/installer/methods/alpine/ 2021-10-25 11:35:19 Cool 2021-10-25 12:29:55 Ariadne thanks for the follow up. Totally understand that moderation isn't 24/7, I really just wanted to make sure that person's attitude wasn't condoned by the project. 2021-10-25 12:30:56 It's certainly not: https://alpinelinux.org/community/code-of-conduct.html 2021-10-25 12:42:42 Hi. I added a bash script in /etc/periodic/hourly but it doesn't get executed. I can execute it manually. In /var/log/messages I see that cron is running: Oct 25 12:00:00 dxsrv07 cron.info crond[2595]: USER root pid 1482 cmd run-parts /etc/periodic/hourly 2021-10-25 12:43:08 does it have +x 2021-10-25 12:45:20 psykose: yes. so I can execute by simply run /etc/periodic/hourly/btrbk.sh 2021-10-25 12:45:50 -rwxr-xr-x 1 root root 97 Oct 25 12:41 /etc/periodic/hourly/btrbk.sh 2021-10-25 12:45:53 could you post the script 2021-10-25 12:46:47 https://dpaste.com/34ZQ62WFV 2021-10-25 12:47:34 well the exec prevents the line after from being run afaik 2021-10-25 12:48:03 unless the exec fails 2021-10-25 12:48:10 (which is different from 'unless btrbk fails') 2021-10-25 12:48:40 hm. But it runs manually. I can skip the exec. It isn't needed 2021-10-25 12:49:26 But why didn't it run anyway? It does when I run from a shell. 2021-10-25 12:50:16 how do you verify that it doesn't run 2021-10-25 12:50:53 also you don't need the sync/sleep in it 2021-10-25 12:51:31 btrbk writes a logfile. But I can add a echo >/tmp/foobar.txt 2021-10-25 12:52:26 I'll add after #!/bin/sh: echo "Running btrbk" > /tmp/cron-test.log 2021-10-25 12:52:37 We will see in 8 minutes =) 2021-10-25 12:52:46 you can use date to force it a bit 2021-10-25 12:53:32 date --set="13:10" etc, then it should trigger a bit sooner :) 2021-10-25 12:53:43 of course you can break things doing that 2021-10-25 12:53:50 and restart ntpd afterward to resync 2021-10-25 12:53:55 Yea,, that I don't want to do 2021-10-25 13:03:42 nope. didn't run 2021-10-25 13:10:01 hmm 2021-10-25 13:11:00 if you run-parts the directory yourself i assume it works? 2021-10-25 13:17:34 zico[m]1, it's not because a mental diseased man entered the room and thrown lots of insults that it makes them "acceptable", the community is small 2021-10-25 13:27:34 jeffreybenzos has been a real disincentive to the community getting any bigger 2021-10-25 13:27:49 i know i peek in here less because of stuff like that. 2021-10-25 13:28:52 but i also know he'll be back in one of probably 100 socks, because that's what irc sociopaths do. 2021-10-25 13:29:27 regarding him specifically i have seen him before over the months, and those times it was minus the trolling as far as my memory goes 2021-10-25 13:29:52 maybe my memory is just not very good currently though 2021-10-25 13:29:52 i think i was around for one of his sessions of trolling 2021-10-25 13:30:08 was really fixated on being an asshole to me specifically at rhe time 2021-10-25 13:30:37 was super weird to me that no one took moderative action esp considering the presence of moderators at the time 2021-10-25 13:30:44 bad memory for me it is 2021-10-25 13:30:49 that sucks to hear 2021-10-25 13:31:13 i dont let internet weirdos get the better of me :p 2021-10-25 13:31:29 its alright 2021-10-25 13:31:35 sucks to hear avout the slurs tho 2021-10-25 13:31:38 me either :p or at least i pretend to 2021-10-25 13:31:55 psykose: run-parts doesn't do anything either. This is odd, isn't it? 2021-10-25 13:32:43 probably missing something extremely obvious 2021-10-25 13:32:43 hm 2021-10-25 13:33:21 As Ariadne pointed out, we should have other ops in this channel that can cover different mimezones. Most of us are in CEST. 2021-10-25 13:33:30 timezones* 2021-10-25 13:33:47 psykose: I noticed something myself as well with bb crond 2021-10-25 13:33:50 oh yea this was in UTC(+1?) late day 2021-10-25 13:34:17 I setup some crons, saw them being exuted in /var/log/messages, but nothing actually being executed 2021-10-25 13:34:29 that's probably the issue here then 2021-10-25 13:34:43 they work fine for me with bb crond, but i wonder what else there is to check here for Forza 2021-10-25 13:34:44 eris[m]: this happened in the middle of the night 2021-10-25 13:34:55 So if anybody here is interested in helping us with this, sent me or ikke a pm 2021-10-25 13:36:01 Saijin_Naib[m]: And I hope you do not feel discouraged. Like others already mentioned, this was not your fault 2021-10-25 13:36:17 I did an strace run-parts: https://dpaste.com/5QPPT72JP 2021-10-25 13:36:33 ikke: that's how it is for me too 2021-10-25 13:37:48 Perhaps run-parts is broken? I could probably just make a crontab instead for each of my maintenance scripts 2021-10-25 13:38:22 strace doesn't show run-parts trying to execute the files it finds 2021-10-25 13:38:35 8:30 AM was super weird to me that no one took moderative action esp considering the presence of moderators at the time 2021-10-25 13:38:44 Forza: runparts is rather picky 2021-10-25 13:38:48 they may not have been paying attention 2021-10-25 13:38:52 why does it return 644 as the mode for the file when it's definitely +x 2021-10-25 13:38:58 unless i don't understand the strace output 2021-10-25 13:39:28 right now all of the moderators we have, are also doing things like sysadmin tasks and dev tasks 2021-10-25 13:41:51 counterpoint: jeffreybenzos is not a new problem 2021-10-25 13:42:10 new to me 2021-10-25 13:42:21 i had not seen him before 2021-10-25 13:43:07 had i seen him before i would have made the mod team aware that we needed to do something about it 2021-10-25 13:43:21 psykose: maybe it is a mask? 2021-10-25 13:43:30 ah perhaps 2021-10-25 13:44:00 if the mod team reads alpine irc channels like i do, they probably do not actively read them except when doing work 2021-10-25 13:44:28 or well, i don't think umask affects existing files 2021-10-25 13:45:11 Ariadne: fair, just pointing out this person's been doing this for a while. i don't log so i can't exactly say when, but i remember months ago him coming in here and complaining about how slow the channel was. 2021-10-25 13:45:24 there's a public log 2021-10-25 13:45:42 https://irclogs.alpinelinux.org/ 2021-10-25 13:46:50 maybe adding a command to algitbot to alert the mods could be helpful 2021-10-25 13:47:20 i mean fundamentally that’s the issue 2021-10-25 13:47:30 we don’t have anyone who is just a mod 2021-10-25 13:47:31 doens't chanserv already have something like that? 2021-10-25 13:47:45 well the idea is to generate a highlight 2021-10-25 13:48:00 so that it’s brought to somebody’s attention that fuckery is afoot 2021-10-25 13:49:11 anyway i need to take a nap and then get a lot of work done so bye 2021-10-25 13:51:06 cheers 2021-10-25 13:57:44 jeffreybenzos doesn't bother me (irc since the 90s, efnet...) so all that's really needed is a +q and point them to -offtopic when it gets noisy 2021-10-25 13:59:37 psykose: no the file was indeed 644. I changed it back to 740. Still no go though 2021-10-25 14:00:13 ah, in your earlier past it was marked rwx 2021-10-25 14:00:46 yes. I know. I have two machines :D 2021-10-25 14:01:02 sorry. but still didnt work 2021-10-25 14:01:22 a command !ops to masshl the OPs could be handy yes 2021-10-25 14:01:31 as long as there are at least one op per timezone 2021-10-25 14:01:40 Forza: it's not allowed to contain periods 2021-10-25 14:01:52 https://github.com/brgl/busybox/blob/master/debianutils/run_parts.c#L115 2021-10-25 14:02:32 the filename to execute 2021-10-25 14:03:44 nice comment on the top 2021-10-25 14:04:31 well that explains it 2021-10-25 14:04:37 * attempt to write a program! :-) . */ 2021-10-25 14:04:46 /* This is my first attempt to write a program in C (well, this is my first 2021-10-25 14:04:46 * attempt to write a program! :-) . */ 2021-10-25 14:04:58 there is always a first :) 2021-10-25 14:05:05 this restriction really doesn't make much sense 2021-10-25 14:05:12 :) 2021-10-25 14:05:23 maybe something with umask? 2021-10-25 14:06:50 this isn't related to umask, it just rejects filenames with a dot 2021-10-25 14:08:11 your file ends with .sh 2021-10-25 14:08:58 psykose: btw, I did not use run-parts 2021-10-25 14:09:20 ie, not /etc/periodic 2021-10-25 14:10:46 psykose: oh. so not .sh? very odd 2021-10-25 14:12:43 like I said, run-parts is picky 2021-10-25 14:13:09 Hi all :-) 2021-10-25 14:14:42 ikke: indeed. run-parts --test /etc/periodic/hourly/ now lists "btrbk" in there 2021-10-25 14:14:53 is there something i can do about lacking freadahead.C? like custom compiler stuff in path or something, im trying to bootstrap something and its all build against gnuC so it expects this for compilation 2021-10-25 14:15:30 I did add edge community repo to my config (tagged as @edge-community), but I'm unable to add firefox-esr@edge-community. It does not seems to take my tag into account while installing firefox@edge-community 2021-10-25 14:15:39 Any idea ? 2021-10-25 14:15:50 did you run apk update afterwards? 2021-10-25 14:16:04 Yes, of course :-) 2021-10-25 14:20:27 How do I list packages availlable per repository ? 2021-10-25 14:27:49 I guess it is not a best practice to install selected package (like firefox requiring many libs) from edge-community 2021-10-25 14:28:03 indeed 2021-10-25 14:28:13 you shouldn't mix stable and edge if that is what you're asking 2021-10-25 14:28:34 Are you using alpine on desktop/workstation ? 2021-10-25 14:29:03 If yes, are you on stable release or edge ? 2021-10-25 14:29:37 kmmndr: apk --repository https://dl-cdn.alpinelinux.org/alpine/edge/main --repositories-file /dev/null search 2021-10-25 14:31:18 clandmeter: Woow nice, thank you very much ! 2021-10-25 15:43:04 For maintaining a package in Community, I looked up what backporting is/entails at a high level, and I think I understand, but if there is a problem with a Community package I maintain that is fixed in a tagged release update, would updating the APKBUILD count as the required "support" for that package? 2021-10-25 15:46:39 Saijin_Naib[m]: typically we only upgrade packages with a certain minor release (there are exceptions) 2021-10-25 15:47:14 if it requires a feature or major update, I'd backport the individual commit instead 2021-10-25 15:49:03 Alright, thank you. If I am concerned about my ability to do such, should I just request it being moved to Testing before it becomes a problem, or cross that bridge if/when it comes? 2021-10-25 15:52:13 the latter, we can help out where needed 2021-10-25 16:04:57 Thank you very much! 2021-10-25 16:21:47 Forza: btrbk fixed that last year in version 0.30.0: https://github.com/digint/btrbk/issues/327 2021-10-25 16:27:04 nickersonm: I see. Thanks. Looking at the man page it does have a "compat busybox" option. 2021-10-25 16:28:18 And indeed I run btrbk on two Alpine installs with no issue and without coreutils. 2021-10-25 16:32:37 for what reason is there gcc6 in the repos? 2021-10-25 16:37:13 For one it's used to compile some kernels in postmarketOS 2021-10-25 16:37:35 interesting 2021-10-25 17:10:00 korner: basically some devices use barely functioning Android kernels that either can't compile with newer kernels or don't work. But there are other cases than this where you may need to use an old compiler, e.g. legacy software 2021-10-25 17:10:51 i know that those devices use those kernels but i didnt expect it being on alpine, only case i can think of where old compiler was needed is palemoon 2021-10-25 20:58:28 PureTryOut (matrix.org): Thank you for providing OSGeo software! I'm trying to package MDAL and GRASS to complement what you've already done 2021-10-25 20:58:44 (and close two early aports issues I raised for wanted packages) 2021-10-26 06:55:36 Saijin_Naib: sorry, OSGeo? What's that? 😅 2021-10-26 09:44:55 Hi. How do I ensure that /etc/profile is sourced when I use screen sessions on Alpine? 2021-10-26 09:46:16 your shell should source that itself, and so then it will source it when screen runs it 2021-10-26 09:47:29 psykose: I'm using the default which is ash. And it does not, at least not for root account. 2021-10-26 09:47:53 you can make a .ashrc with a source in it 2021-10-26 09:47:54 do I need a ~/.bashrc or similar? 2021-10-26 09:47:57 ah :) 2021-10-26 09:48:00 of course 2021-10-26 09:48:28 nope =( 2021-10-26 09:48:40 actually no, hah 2021-10-26 09:48:42 maybe screen needs to do a login session or something 2021-10-26 09:49:29 i think if you have it as a login shell it sources .profile, where you can source /etc/profile 2021-10-26 09:49:33 let me go read the code real quick 2021-10-26 09:49:53 Aha. Thanks. No hurry! =) 2021-10-26 09:49:56 brb 2021-10-26 09:54:14 ash -l seems to source /etc/profile itself, and ~/.profile 2021-10-26 10:14:08 It's only when I start screen. normal ssh login works and it sources the profile 2021-10-26 10:16:51 I needed ~/.screenrc 2021-10-26 10:16:54 :) thanks again 2021-10-26 14:51:14 PureTryOut (matrix.org): I thought you maintained some OSGeo packages like GDAL, PROJ, etc. It is enitrely possible it was late and I got things mixed. Regardless, it looks like you provide the entire KDE/QT stack, so thanks for that then 🤣 2021-10-26 14:54:21 what is program on alpine to use to extract rar archives? seems like neither urar or unrar are available 2021-10-26 14:55:53 unarchiver, maybe? i dunno if that's in there 2021-10-26 14:56:05 hi 2021-10-26 14:56:16 no, it's not. 2021-10-26 14:56:17 ACTION sent a code block: https://matrix.org/_matrix/media/r0/download/matrix.org/QGyyiglvEweHRjBAAUGVpnbZ 2021-10-26 14:56:29 where? i apk cant seem to find it 2021-10-26 14:56:40 unrar is available 2021-10-26 14:57:00 apk search unrar 2021-10-26 14:57:01 i have main community and testing enabled, using edge as well, i dont know where else could it be 2021-10-26 14:57:53 it is in non-free 2021-10-26 14:58:02 right 2021-10-26 14:58:36 moved from main 2 weeks ago 2021-10-26 14:58:46 thats community repo, and i do have it enabled, how odd 2021-10-26 15:00:08 you can also try p7zip 2021-10-26 15:00:23 rar isnt zip 2021-10-26 15:01:42 https://documentation.help/7-Zip/formats.htm 2021-10-26 15:01:50 Indeed not, but p7zip handles a lot of compression formats 2021-10-26 15:02:27 indeed it is not, yet p7zip extracts rars 2021-10-26 15:02:53 i wish midnight commander would see it, guess i gotta extract it manually 2021-10-26 15:04:02 hmmmm i get error of unknown suffix, maybe this p7zip package is built without rar feature? 2021-10-26 15:06:38 7zr x file.rar? 2021-10-26 15:06:55 yeah i found that just now, i assumed the command is p7zip 2021-10-26 15:19:16 Saijin_Naib: I did package and now maintain the KDE stack yes, and I maintain the Qt stack indeed. But I never heard of OSGeo and I don't maintain the packages you mentioned 2021-10-26 15:47:15 hallo 2021-10-26 16:40:03 Hello everyone. What is the expected batterylife if running alpine without windowmanager or anything? Acpi and powertop tell me that with my brand new battery of 4400mAh on full charge i have only about 7hours, is that low? 2021-10-26 16:40:55 If it is too low, what can I do to make it go longer? I just want to use tmux, gcc, python3, git and vim so I dont need much to run 2021-10-26 16:41:28 so you want to develop and compile, things that are cpu intensive 2021-10-26 16:42:19 without those variables, it would depend on your hardware 2021-10-26 16:42:48 but acpi tells me only that only seven hours are remaining after a fresh start with nothing running 2021-10-26 16:43:06 maybe that's good 2021-10-26 16:43:20 is the screen on? 2021-10-26 16:43:51 yes my screen shows the tty 2021-10-26 16:45:52 So if this is normal and there is nothing i can config in alpine to improve the battery life I am fine. 2021-10-26 16:45:56 Thank you :) 2021-10-26 16:47:00 for a while i ran a setup like that with distcc/et al offloading building to another machine (threadripper) that'd i'd bring up with wake-on-lan 2021-10-26 16:47:29 so that it was basically a low-power dumb terminal 2021-10-26 16:47:49 you can do a lot, just depends on how creative you want to get with it 2021-10-26 16:48:55 Ah, i see. Thanks. 2021-10-26 16:50:55 I was trying to deactivate wireless connections as i dont need them but did not find any manual, can you link to one? I am happy to read and configure myself just need help locating the right manual or finding the right tools 2021-10-26 16:53:27 best done from bios/efi 2021-10-26 16:55:27 if the machine has no hardware or bios switch for it, rfkill may do it 2021-10-26 16:56:26 thanks, my bios does not have anything about wireless will read rfkill manpage 2021-10-26 16:59:41 I used rfkill to block everything but it does not increase battery life 2021-10-26 16:59:43 the backlight is still technically at whatever high value it can be, wm or not 2021-10-26 16:59:51 you could try lowering that too 2021-10-26 16:59:57 but thank you very much 2021-10-26 17:00:34 tbh 7 hours isn't bad 2021-10-26 17:03:33 thanks, I think I will start playing with the backlight and then leave it as it is 2021-10-26 17:05:54 there's a guy who swapped out the display on a thinkpad with eink 2021-10-26 17:06:06 i've been wanting to do something like that for a while. 2021-10-26 17:06:10 what kind of computer is more useful than how many mAh (which is usually a lie anyways) 2021-10-26 17:06:17 one sec, let me find the url 2021-10-26 17:06:19 invoked: Do you have a link to that mod? 2021-10-26 17:06:37 https://alexsoto.dev/building-an-e-ink-laptop.html 2021-10-26 17:06:40 mAh is only helpful if you know how many A something uses. :) 2021-10-26 17:08:08 he doesn't really say what that does to battery life though. :( 2021-10-26 17:08:12 i'd assume it's a win 2021-10-26 17:08:51 sans backlight would probably be an improvement 2021-10-26 17:09:34 i like the idea i'd be able to use a laptop in full sun, also 2021-10-26 17:09:51 it'd be interesting for a while, anyway. 2021-10-26 17:11:09 getting a remarkable tablet is something i'm eventually going to do. i know those are a little hackable. 2021-10-26 17:17:48 I would love to have an eink display but they are really expensive and mostly proprietary 2021-10-26 17:18:20 My machine has 1GB of RAM and an 1.5Ghz Intel Atom (it is 9years old) 2021-10-26 17:18:47 the only new parts are a 500GB SSD and the new battery 2021-10-26 17:19:53 https://www.youtube.com/watch?v=tc1Fzj46voc it's nice how it's not ghosting like 1st gen eink 2021-10-26 17:20:08 not that i'd play video games on it, but, it'd be a stable work display. 2021-10-26 17:22:34 maybe MNT will introduce an eink option someday. 2021-10-26 17:22:35 thats stunning. For tty work it is almost overpowered 2021-10-26 17:25:28 i actually own a e-ink reader. It is so easy on the eyes, but again they are sadly mostly expensive and proprietary 2021-10-26 17:25:58 the way things are going, eink is pretty niche and will probably remain that way outside of things like remarkable. 2021-10-26 17:26:05 but i can hope 2021-10-26 17:26:16 an org like MNT is capable of doing it someday. 2021-10-26 17:26:45 colour eink is even more niche 2021-10-26 17:28:33 stupid question: what is mnt? 2021-10-26 17:28:48 https://mntre.com/#about-us 2021-10-26 17:28:55 not a stupid question 2021-10-26 17:29:03 If MNT published the Reform's schematics someone could DIY one without the usual reverse engineering effort. 2021-10-26 17:29:17 I think they do 2021-10-26 17:29:45 ACTION wants one but has plenty of used laptops that still work... 2021-10-26 17:29:59 getting things manufactured is the real challenge, MNT understands this 2021-10-26 17:30:25 i don't think they're bothered if someone else tries to go their own way to get a reform built 2021-10-26 17:30:31 because it's hard 2021-10-26 17:30:33 ah thank you, thats really interesting 2021-10-26 17:32:37 ARF: i've been wanting a reform for a while but they're always on backorder 2021-10-26 17:32:52 for some reason i just can't buy things on backorder. one of my hangups 2021-10-26 17:33:37 same goes for pine 2021-10-26 17:33:45 i'd have them already if they just had them in stock, ever. 2021-10-26 17:34:52 I have a pinephone with PostmarketOS as a backup phone. It's actually what got me using Alpine 2021-10-26 17:35:18 I wanted to get wsjtx to work on it and doing it all on a phone was kind of hell. 2021-10-26 17:35:42 how do you feel about the pinephone, overall, now? 2021-10-26 17:36:48 ACTION apologizes for being offtopic 2021-10-26 17:40:10 It's cheap and slow, but that's not unexpected. 2021-10-26 17:42:43 yeah. and what i want to do with a phone is so basic computationally, that i'd probably main the pinephone once i get one. 2021-10-26 17:43:58 honestly i just want a phone-like device without the telephony at all, just data. 2021-10-26 17:44:45 The hardware works fine. Software has a lot of rough edges. It gets on WiFi, mobile data, sends SMS and makes phone calls just fine though. 2021-10-26 17:46:05 I think I have an issue with iwlwifi 2021-10-26 17:46:15 after suspend/resume wifi ssid stays the same but no longer works 2021-10-26 17:47:03 invoked: I think that's called a tablet :-) 2021-10-26 17:47:30 probably something with your specific adapter. you're not running elogind and acpi at the same time are you 2021-10-26 17:48:16 minimal: but i want it in my pants pocket, man 2021-10-26 17:49:52 modern phones are creeping up to the size of small (i.e. 7") tablets 2021-10-26 17:49:55 It is probably cheaper to buy pants with a bigger pocket. 2021-10-26 17:50:35 i don't want to dress like mc hammer 2021-10-26 17:51:20 :) 2021-10-26 17:51:41 thought the 80s/90s look as back in fashion lol 2021-10-26 17:51:46 s/as/was/ 2021-10-26 17:52:13 complete with "brick" mobile phones 2021-10-26 17:53:29 the new thing will be everyone strapping a computer to their faces, en masse 2021-10-26 17:54:21 i'm fine with having a small device in an average-sized pants pocket, and testicular cancer 2021-10-26 18:01:44 "My machine has 1GB of RAM and an..." <- Stand-by CPU usage from an old Atom like that is likely pretty high. have you checked PowerTop to see what everything is drawing? 2021-10-26 18:03:26 someuser: For context, my N3450 4GB RAM 64GB eMMC system with around 34.2WH battery gives me roughly 13hr in terminal at 5% brightness (honestly more than enough for night into typical lit rooms) 2021-10-26 18:04:07 Once I start browsing or using GNOME Builder to compile documents, that can tank pretty quickly as the TDP on this chip can exceed 6W, taking that battery down to maybe 4.5hrs of use 2021-10-26 18:04:38 Yeah that's how battery power works. 2021-10-26 18:05:40 I will take my time and analyse what powertop says 2021-10-26 18:06:08 Yeah, I'm aware, but the difference is quite shocking. On an Atom chip of my wife's laptop from around the same generation as @someuser, the passive power isn't much less than active, unfortunately. She gets about 3hrs no matter what she's doing 2021-10-26 18:06:41 someuser: you could also check out tlp and maybe forcing the intel pstate driver in powersave mode for the CPU governor. I'm using those myself 2021-10-26 18:08:17 its not shocking at all when you understand all the things that modern computers do to save power mostly boil down to turning things off when they aren't being used. 2021-10-26 18:10:52 Yep 2021-10-26 18:10:57 i've long wanted a laptop that had two hot-swappable batteries. you can just swap one out as you need to and keep running, and keep the spares on chargers 2021-10-26 18:11:15 that would be ideal, imo 2021-10-26 18:11:38 You need a time machine to go back to the late 90's :) 2021-10-26 18:11:40 if you are near chargers for spares you can just plug in the laptop though 2021-10-26 18:11:42 invoked: The closest thing I had was my EliteBook 2740p, but only the flat battery could be hot-swapped, the internal had to be done with AC in 2021-10-26 18:12:07 unless you mean hot swappable without powering off and having a bunch of them 2021-10-26 18:12:12 psykose: yeah but on a typical day i'm grabbing my laptop and going from meeting to meeting 2021-10-26 18:12:16 my life sucks, in other words 2021-10-26 18:12:19 Even with both packs and something like 90WH of battery, if maybe could get 4hrs 2021-10-26 18:12:20 haha 2021-10-26 18:12:25 what is tlp? I am new to this hardware stuff 2021-10-26 18:12:47 https://pkgs.alpinelinux.org/package/edge/community/x86_64/tlp 2021-10-26 18:13:19 Sane defaults, but you can get some extra mileage out of tweaking the config file. Incredibly well documented, IMO, so pretty easy to read and understand what to change and how 2021-10-26 18:13:50 I seem to remember a Dell laptop about 14 years ago where the batteries and CDROM removable units were the same size so you could have 2 batteries fitted 2021-10-26 18:14:14 That sounds like an Inspiron, yeah. Haha. Tank machine 2021-10-26 18:14:29 Did the ThinkPads ever come with add-on batteries? That could be a nice unit 2021-10-26 18:14:56 Yeah they did 2021-10-26 18:15:04 they came with bigger battery options, still do, probably 2021-10-26 18:15:09 ones that would stick out 2021-10-26 18:15:12 Some units (like my T60) could have a 2nd battery instead of the CD-ROM drive. 2021-10-26 18:15:37 Others (like my T450) have an internal Li-Po cell and a 2nd removable one. 2021-10-26 18:15:58 I think they also made some that attached to the bottom via the docking connector. 2021-10-26 18:16:04 that would probably work for me. i probably have one of those in my laptop stack somewhere, i just never bothered to use it that way. 2021-10-26 18:16:20 Oh, that's awesome. Did they have "fuel-gauge" batteries like the EliteBooks did? That was really nice to be able to tap the LED strip to get a charge level, especially when you have tons of extras around 2021-10-26 18:18:20 ah, thank you very much 2021-10-26 18:19:26 someuser: I also am curious if the linux-edge kernel might do power management better than the linux-lts kernel, but powertop will not work under linux-edge since it is missing a module (I think debugfs, it was). I've not spent the time to profile it via wall clock yet 2021-10-26 21:32:27 ls 2021-10-26 21:33:09 quit 2021-10-26 22:38:48 guys, I remember there was a group plugdev, that allowed usb device acceess. Can't find it anymore 2021-10-26 22:39:02 I need to setup u2f for forefox - how do I do it now? 2021-10-26 23:44:47 I'm trying to make a package of a program which requires dropping a directory of the compiled language runtime VM (Erlang) and internal libs onto the filesystem somewhere. I would usually just plop the directory in /opt, but abuild tells me that is not allowed. Where is an appropriate place in the filesystem for me to put this directory? 2021-10-26 23:46:06 /usr/lib probably? 2021-10-26 23:46:35 is there no way to use the system erlang as well 2021-10-26 23:46:39 that would make it a bit more obious 2021-10-26 23:49:26 as far as I can figure out, the program still needs to use a directory with internal deps bundled in even if erlang is installed on the system. And I would rather not package up 100+ insternal dependencies if I can get around it 2021-10-26 23:50:05 i don't know what this package is or what the directory structure looks like, of course you would keep the libs in it 2021-10-26 23:51:36 ok, /usr/lib sounds like the right place, thanks. The directory will include the binaries for the program, that's why I wasn't sure 2021-10-26 23:52:46 and that this is the first time I'm packaging something, so its a lot of banging my head against a wall until I understand how to do it 2021-10-26 23:53:22 mhm 2021-10-27 00:39:37 weird; firefox on edge is using alsa as its audio backend instead of "pulse" (through pipewire-pulse 2021-10-27 00:39:39 ) 2021-10-27 00:39:55 output works fine, but my mic isn't read 2021-10-27 00:41:04 something tells me it's within how firefox handles that build flag we use to enable alsa-only 2021-10-27 00:47:08 it doesn't enable alsa only, they're both enabled 2021-10-27 00:47:53 psykose: yeah, though i'm surprised it defaults to alsa first 2021-10-27 00:48:03 i wonder if something's fucky with my pipewire config now that it supposedly started up 2021-10-27 00:48:15 it doesn't, it uses pulse for me with both available 2021-10-27 00:48:17 seeing as pipewire-media-session isn't running...? 2021-10-27 00:48:20 there's probably some other weird issue 2021-10-27 00:48:32 without media session you can't output anything to pipewire i don't think 2021-10-27 00:48:46 that's what i thought, so idk why my config is incorrect now 2021-10-27 00:49:06 how do you start pipewire 2021-10-27 00:49:13 currently running it on xinit through pipewire &>~/pipewire.err & 2021-10-27 00:49:35 that hasn't worked for a while, the default config doesn't start pipewire-pulse or the session manager 2021-10-27 00:49:58 unless you changed it of course 2021-10-27 00:50:00 lul, nice on my part 2021-10-27 00:50:09 i don't think i did, but i thought it already started that 2021-10-27 00:50:15 in any case, copy /usr/share/pipewire/pipewire.conf to /etc/pipewire/pipewire.conf 2021-10-27 00:50:17 what's the modern (as of weeks ago) method of launching it? 2021-10-27 00:50:18 ah 2021-10-27 00:50:30 at the bottom you will see two lines, one for pipewire-pulse and one for media-session, just uncomment them 2021-10-27 00:50:37 you can also install wireplumber instead of pipewire-media-session 2021-10-27 00:50:44 and replace pipewire-media-session with wireplumber in there 2021-10-27 00:51:01 newer session manager, media-session will eventually be deprecated 2021-10-27 00:51:03 What's wireplumber? 2021-10-27 00:51:07 Mmm, huh. 2021-10-27 00:51:09 same thing but different 2021-10-27 00:51:33 if you never configured anything the defaults should still work, bluetooth works a bit better with plumber 2021-10-27 00:51:45 up to you though :) just those two lines in config uncommented and restarting should work 2021-10-27 00:52:42 pipewire also does ship with a /usr/libexec file that executes the things for you instead of using the file, but for some reason it randomly bugged a little starting it sometimes like that, better to just do it in the actual config like here 2021-10-27 00:54:48 what's the easy way to restart pipewire without killing my whole session? 2021-10-27 00:55:08 is it enough to just kill that first process and then run that line myself somewhere? 2021-10-27 00:56:02 pkill pipewire then run it again 2021-10-27 00:56:25 it doesn't touch anything else, or at least i do that all the time when testing various things 2021-10-27 00:57:08 just realized i'm running this window in firefox lmao 2021-10-27 00:57:12 brb i guess 2021-10-27 00:57:24 you don't have to restart firefox i don't think 2021-10-27 00:57:43 at least it always lets me play audio even through pipewire restarts, but maybe with alsa that doesn't work 2021-10-27 01:00:33 firefox audio is broken again, lmao 2021-10-27 01:00:36 no output 2021-10-27 01:01:12 no input either, as tested by me going to some mic test site that shows your waveform 2021-10-27 01:01:24 firefox has chosen pulse-rust as the backend 2021-10-27 01:02:21 what is pulse-rust 2021-10-27 01:02:25 yeah, dunno 2021-10-27 01:02:41 apk info | rg rust yields nothing, of course 2021-10-27 01:02:46 if you run pactl info what do you see 2021-10-27 01:03:02 i don't have pactl installed 2021-10-27 01:03:20 pulseaudio-utils 2021-10-27 01:03:27 leery eyes 2021-10-27 01:03:35 what would i need pulseaudio-utils for? 2021-10-27 01:04:05 to run pactl info 2021-10-27 01:04:18 sounds like a clever trick to get me to run poetteringware 2021-10-27 01:04:42 ok 2021-10-27 01:04:46 Server Name: PulseAudio (on PipeWire 0.3.39) 2021-10-27 01:04:55 Default Sink: alsa_output.pci-0000_01_00.1.hdmi-stereo 2021-10-27 01:04:57 found out why 2021-10-27 01:05:00 dunno why that's my sink 2021-10-27 01:05:25 if i understand "sinks" correctly, the thing i want audio to go to is what is currently the "source" 2021-10-27 01:05:31 Default Source: alsa_input.usb-C-Media_Electronics_Inc._USB_Audio_Device-00.mono-fallback 2021-10-27 01:06:00 wouldn't this just route all my output from my intended audio card (usb audio) to my monitor's output? 2021-10-27 01:08:14 those are correct sinks i.e. the sink is the output, and goes to hdmi stereo 2021-10-27 01:08:27 if you want to see if they're set correctly you need to use something that configures the outputs 2021-10-27 01:08:46 first check if you have everything running, ps aux | grep wire should give you the 3 things 2021-10-27 01:09:02 then you can use pulsemixer or pavucontrol to configure things like back in pulseaudio land 2021-10-27 01:09:09 there's no pipewire-native config tools yet that i know of 2021-10-27 01:10:48 also check ps for grep pulse to see if you had another mystery pulseaudio this whole time 2021-10-27 01:10:56 psykose: i've never been "in pulseaudio land" 2021-10-27 01:11:04 i've been in alsa land and alsamixer works fine for mixing 2021-10-27 01:11:15 and chooses the proper default card as i've configured it 2021-10-27 01:11:45 alsamixer doesn't see anything from pipewire for me 2021-10-27 01:12:42 i'm merely giving you a list of things to potentially do to see that everything is in place 2021-10-27 01:12:55 aye; i see i see 2021-10-27 01:13:16 what's pipewire-alsa for? 2021-10-27 01:13:49 it routes alsa connections to pipewire 2021-10-27 01:14:15 or however one correctly says that in audio terms 2021-10-27 01:15:08 if you're using media session you want a with-alsa and with-pulseaudio file in the media-session.d directory if i remember correctly, otherwise they're there by default in wireplumber iirc 2021-10-27 01:16:07 is wireplumber supposed to be at the root of all things if it's going to be used? 2021-10-27 01:16:22 for example, would i launch everything by launching wireplumber? 2021-10-27 01:16:32 no, pipewire launches the other things 2021-10-27 01:16:37 ah, interesting 2021-10-27 01:16:54 i asked that specifically because it said pipewire-media-session appears to be running, stop it before starting wireplumber 2021-10-27 01:16:54 the two things in the config file are basically just an exec of them 2021-10-27 01:16:59 yep 2021-10-27 01:17:07 i just monkey'd in wireplumber to see if it was one command fix all 2021-10-27 01:17:09 if you kill pipewire the other die if they were started by it 2021-10-27 01:17:32 something tells me the wiki page is out of date 2021-10-27 01:17:48 which one 2021-10-27 01:18:30 the one for pipewire 2021-10-27 01:19:22 i thought with-pulseaudio implied that pulse was actually installed on the system? 2021-10-27 01:19:40 no, pipewire is the pulse daemon 2021-10-27 01:19:49 actual pulseaudio doesn't have to be there 2021-10-27 01:20:12 and the wiki page is accurate for pipewire+media-session, based on a glance 2021-10-27 01:22:12 i've got with-alsa in there 2021-10-27 01:22:37 what's the method of configuring pipewire without pulseaudio tools? it's always been drop-in working for me 2021-10-27 01:23:09 afaik there isn't one 2021-10-27 01:23:31 poettering trying to win yet again 2021-10-27 01:23:31 or well, you can use the pipewire-utils things 2021-10-27 01:23:45 though you will be doing a lot of jq parsing and passing numbers around to set profiles 2021-10-27 01:23:47 what i don't get is how i'm going to define a source at all when it's not an audio card i need to source from 2021-10-27 01:23:53 i just want the audio out of my usb card 2021-10-27 01:24:16 i would think and hope that pipewire would just read my alsa config and use that 2021-10-27 01:24:34 no, it's its own device 2021-10-27 01:24:38 what do you set for the device 2021-10-27 01:25:12 what have i set? i've set kernel params so that it's the first card and then set alsa's config to use it 2021-10-27 01:25:28 that just sounds like picking a default output 2021-10-27 01:25:51 right, so why isn't pipewire picking the same 2021-10-27 01:25:54 which you can do in a few seconds in any of the pulse guis 2021-10-27 01:25:59 or clis 2021-10-27 01:26:31 why would the pulse stuff inside of pipewire not already have what i've got as the default? it makes me think i've got default issues somewhere else 2021-10-27 01:26:32 it's not picking it because it's not alsa and not reading the alsa config, after you pick it the session manager state should remember it 2021-10-27 01:26:34 or misconfigured something 2021-10-27 01:26:59 so what's the default mode of picking it? pavucontrol wasn't in pulseaudio-utils 2021-10-27 01:27:07 pavucontrol is pavucontrol 2021-10-27 01:27:16 pulsemixer is pulsemixer 2021-10-27 01:27:21 i use the latter, either have the same options 2021-10-27 01:27:46 ah, perhaps this is why 2021-10-27 01:27:56 i think pavucontrol or whatever behind it thinks that my card is input only 2021-10-27 01:28:19 you can see under configuration the profiles for it 2021-10-27 01:29:04 i see mono input and pro audio 2021-10-27 01:29:17 no wonder people can't stand pulse 2021-10-27 01:29:45 how do you set it to both in alsa 2021-10-27 01:29:47 we haven't had a single stage along this journey that wasn't immediately blocked by some other random obstacle 2021-10-27 01:29:56 you don't; it just works, lmao 2021-10-27 01:30:06 poetteringware in shambles yet again 2021-10-27 01:32:53 firefox not playing audio is gonna kill me 2021-10-27 01:33:04 i'm gonna kill pipewire-pulse in the config to see if i can recover 2021-10-27 01:34:42 if pipewire-alsa exists and it goes through that it ends up at pipewire the same 2021-10-27 01:34:43 brb 2021-10-27 01:34:57 i don't have pipewire-alsa right now, though i'll talk about that soon 2021-10-27 01:36:06 To clarify my usecase, I've got chromium for discord and do everything else in firefox without localstorage enabled (never remember history) 2021-10-27 01:36:34 i've got polybar with some alsa stuff for scrolling on a bar to control volume, and sxhkd has some amixer calls as well 2021-10-27 01:37:06 i added pipewire in hopes of fixing my mic input on chromium and that seemed to work just fine, though it doesn't work on firefox, and that was fine enough 2021-10-27 01:37:40 doing most simultaneously input-output stuff was fine on chromium, but when i recently upgraded, having pipewire-alsa installed (no clue what i'm doing) seemed to replace the default card with "pipewire" 2021-10-27 01:37:50 no clue if that created a loop or not 2021-10-27 01:37:53 but no sound was working at all 2021-10-27 01:38:04 the issue here is the profile on the card 2021-10-27 01:38:08 on alsa it was whatever correct profile 2021-10-27 01:38:20 now it's either input or 'pro audio', which is probably because of.. idk 2021-10-27 01:38:27 maybe some card quirks that are handled in alsa only 2021-10-27 01:38:31 so i uninstalled pipewire-alsa and got sound in chromium working fine whereas stuff wasn't on pulse in firefox anymore, just using alsa as backend, so the mic input doesn't work again, but my audio output does 2021-10-27 01:38:36 what you can try is set pro audio, then you will get a million input/outputs 2021-10-27 01:38:50 one of them somewhere probably works for both 2021-10-27 01:38:50 i tried setting pro audio and got absolutely no change in the output area 2021-10-27 01:38:55 ah 2021-10-27 01:39:11 that info is at least helpful toward whatever i'm gonna end up doing 2021-10-27 01:40:04 well not having pipewire-alsa is running two stacks, and the things are just accessing alsa directly themselves, and pipewire isn't really doing anything here 2021-10-27 01:40:21 so something in pipewire doesn't configure your card correctly 2021-10-27 01:40:43 you can try ask in #pipewire (though it's late at night) or open an issue with a bunch of debug info on the pipewire project if you care enough to spend time on it 2021-10-27 01:41:03 but if pro audio also doesn't show anything (and you didn't miss something obvious in these steps that i can't guess) then there's not much you can do 2021-10-27 01:41:25 ideally you have pipewire-* running for all the audio to go through it, but the output just doesn't work here 2021-10-27 01:41:35 that's what i'd figure. i assume that if it detects more than one card, pavucontrol lets you choose from the dropdown by clicking the device 2021-10-27 01:42:15 under profile: xyz in configure you choose the profiles yes 2021-10-27 08:28:00 In a chroot with QEMU s390x user emulation, I can't increase the stack size. ulimit -Hs shows unlimited and ulimit -Ss shows 8192. But after running ulimit -Ss 16384, ulimit -Ss still is at 8192. This doesn't happen outside of the QEMU chroot. Also setting (for example) core file size with ulimit -c 1024 works 2021-10-27 14:26:54 anyone using usb ethernet adapter on raspberry pi? and know what to do (or which extra modules need to add) to start right after boot? 2021-10-27 14:28:05 in my case popular UE300 from tp-link based on rtl8153 which using CONFIG_USB_RTL8152=y in kernel 2021-10-27 14:28:53 but it doesnt work at all until I plug it out and back in again 2021-10-27 14:32:37 lets see if extra firmware will change that behavior... brb 2021-10-27 14:37:08 nope so I bet there is some module which have to be loaded early enough... 2021-10-27 14:39:52 does your rc-update show hwdrivers in the default runlevel? 2021-10-27 14:41:02 Saijin_Naib[m]: no 2021-10-27 14:42:50 Saijin_Naib[m]: wondering if that will fix it, I will check, thank you! still why it isnt default or never seen it before :\ 2021-10-27 14:44:01 reboot and we will see 2021-10-27 14:50:07 Saijin_Naib[m]: nope, didn't help 2021-10-27 14:50:16 Ah, no fun 2021-10-27 14:51:25 after replugin usb adapter I see lsmod got extra 2 modules loaded, will try load them as early possible 2021-10-27 14:52:33 MY-R: you said it works when you remove it and plug back in? why no do "lsmod > mods-before.out", unplug and replug is, do "lsmod > mods-after.out" and diff the 2 files to see if some additional module is being loaded? also see if any different appears in "dmesg" output at boot versus when you plug it back in again 2021-10-27 14:53:00 minimal: ye I did it :) 2021-10-27 14:53:12 so add those 2 to /etc/modules file perhaps? 2021-10-27 14:53:23 ye going to reboot and will see 2021-10-27 14:53:40 lucky me it is my router and irc... 2021-10-27 15:01:31 minimal: nope, didnt change anything, eh Im rly tired of that.... 2021-10-27 15:02:19 so after boot "lsmod" didn't show the 2 modules loaded? 2021-10-27 15:02:41 thought using udev (eudev) will save me from another weird things like that but looks like nope 2021-10-27 15:02:46 is this run-from-ram or "sys" (disk install)? 2021-10-27 15:03:25 minimal: it loaded those modules but usb adapter wasnt detected at all and module for it is compiled in kernel 2021-10-27 15:03:36 minimal: disk install 2021-10-27 15:04:04 working fine from usb-sata disk adapter 2021-10-27 15:04:57 wondering if the modules are loaded "too late", i.e. after wireless was initialised? 2021-10-27 15:05:29 could try adding these modules to initramfs to get them loaded earlier 2021-10-27 15:05:51 what are the names of the 2 modules? 2021-10-27 15:06:08 sha256_generic and cdc_ether 2021-10-27 15:06:46 it is usb-ethernet gigabit adapter, network is loaded after modules service 2021-10-27 15:07:59 try adding these to /etc/mkinitfs/features.d/network.modules and rebuild initramfs? 2021-10-27 15:08:15 oki 2021-10-27 15:12:34 ok, added kernel/drivers/net/usb/cdc_ether.ko and kernel/crypto/sha256_generic.ko mkinitfs; reboot 2021-10-27 15:13:03 if you do "modinfo cdc_ether" perhaps it declares a dep on sha256_generic 2021-10-27 15:16:25 minimal: nope, same 2021-10-27 15:16:59 Im kind of sure that r8152 shouldnt be compiled in to kernel and would work just fine and load fine as module 2021-10-27 15:18:42 but like I said Im kind of tired to fight again with something like that on Alpine, and wont compile kernel for another 6 hours just to check it until wont get something to cool down my raspberry to decent temperature 2021-10-27 15:19:53 and for package with cool box for rpi4 will have maybe after 2 weeks.... (damn delays :/ ) 2021-10-27 15:19:56 guessing either (a) the wifi driver should be a loadable module or else (b) these 2 modules need to be compiled in 2021-10-27 15:20:33 ye 2021-10-27 15:21:38 MY-R: I have built a aarch64 VM that I'm intending to use with QEMU for building RPI kernels 2021-10-27 15:23:46 minimal: I was trying that and on my i5 4460 was slower than on rpi4 which was all time throttling hard 2021-10-27 15:24:40 not to mention APKBUILD which have to build rpi and rpi4 kernel so double time and need all time trim APKBUILD to avoid build also dev package which taking time too 2021-10-27 15:24:57 PC might be slower but I'd guess better cooling 2021-10-27 15:26:01 ordered some passive box for rpi which can connect to 5 chips on rpi4 board, then can add some fan or stick to it extra big radiator 2021-10-27 15:27:28 but ye those small radiators on steel box arent nice, could add fan but gonna replace it soon so wont play with it, I thought that will play with rpi with extra network adapter instead of vlans 2021-10-27 15:28:52 what's the name of the wifi driver/module? 2021-10-27 15:29:27 I was so happy that plugged in adapter and boom working right after but then sad as F when seen that it didnt load and that is Alpine thing, not first time 2021-10-27 15:30:21 minimal: rtl8153 usb3 to ethernet gigabit dapter (UE300 TP-Link) popular and well known adapter 2021-10-27 15:46:06 MY-R: strange rtl8153 and cdc_ether are effectively "conflicting" drivers 2021-10-27 15:46:46 are you sure rtl8153 is the right driver for your adaptor? some vendors change chipset in revisions of adaptors 2021-10-27 15:48:22 cdc_ether is a "industry standard" for some USB network adaptors 2021-10-27 15:53:59 I just have to say I <3 how abuild installs build dependencies it needs then removes them when its done. 2021-10-27 15:54:42 ARF: So well thought out 2021-10-27 16:27:33 MY-R_: have you the linux-firmware-rtl_nic package installed? Seems like its needed for UE300 adaptor 2021-10-27 16:46:01 minimal: wasnt needed, was working without too (of course after repluging) but installed it too and same behaviour 2021-10-27 16:48:13 MY-R_: my point is that the rtl8153 driver seems to need firmware - if the driver is build into the kernel then how will it get firmware when it initialises? If you plug it in after the machine is booted then it *can* load the firmware as the rootfs is already mounted including the firmware in /lib/firmware/rtl_nic/ 2021-10-27 16:50:06 minimal: ye but even without firmware that driver was working, saw only error that couldnt load firmware but still was working 2021-10-27 16:50:50 MY-R_: doesn't the "couldn't load firmware" message mean it is *NOT* working as it needs firmware... 2021-10-27 16:51:31 probably was using some generic stuff but was working, maybe just didnt have extra features but routing/nat etc was fine without firmware 2021-10-27 16:52:15 MY-R_: it would be unusual for a device to partly work if it needs firmware but none is found, it would tend to not work are all 2021-10-27 16:52:21 then thought that I need firmware but it is something in Alpine boot process or just shouldnt be that module compiled in 2021-10-27 16:52:35 firmware is not a kernel module, its a file 2021-10-27 16:53:12 nothing unusual, saw it with different nics or even with radeon stuff 2021-10-27 16:53:30 the driver code appears to have a retry for loading firmware so perhaps adding it to the initramfs (used in early boot) might get it loaded 2021-10-27 16:55:22 adding "/usr/firmware/rtl_nic" to /etc/mkinitfs/features.d/network.files and rebuilding the initramfs 2021-10-27 16:57:03 /lib/firmware/rtl_nic, will try but Im sick of those reboots 2021-10-27 16:57:52 the cdc_ether being loaded appears to be due to udev loaded it when the adaptor is plugged in as it "looks" like a cdc_ether device - but the cdc_ether module actually has a built-in blacklist for "flakey" devices including the UE300 so it ignores it (and so the r8152 driver is then able to use it 2021-10-27 17:00:46 minimal: added all, reboot 2021-10-27 17:17:52 minimal: looks like drivers are loaded at th begining and dont need add firmware or any module but interface isnt created at all cus cant find device eth1, after repluging device is visible 2021-10-27 17:17:55 blehh 2021-10-27 17:19:05 does the "dmesg" still show firmware loading error during boot? 2021-10-27 17:20:02 minimal: since installed firmware it doesnt show it, even if removed firmware dir from initramfs 2021-10-27 17:20:57 ok, so the retry *does* wait long enough for the rootfs to be mounted (and so no need to add firmware into initramfs) 2021-10-27 17:21:32 so what dmesg output does appear during boot regarding the device (before you unplug/replug)? 2021-10-27 17:22:08 minimal: the only diffrence is that: https://tpaste.us/plRw 2021-10-27 17:24:31 maybe that driver should be added after udev start so could create eth1 somehow 2021-10-27 17:24:42 My-R: is this one of those USB devices that can appear as different devices at different times (some USB 3G/4G adaptors do that) where for example it looks like a storage device initially (so Windows/Mac can get the drivers from the storage drive)? 2021-10-27 17:25:09 in dmesg after every reboot it looks all same 2021-10-27 17:25:10 or is the "usb-storage 2-1:1.0: USB Mass Storage device detected" dmesg output related to a different device? 2021-10-27 17:25:46 is always same address of usb 2-1 2021-10-27 17:26:19 I mean do you have a USB storage device also attached to the RPI as well as the UE300? 2021-10-27 17:26:43 Alpine boot from USB to SATA adapter 2021-10-27 17:28:27 ah ok 2021-10-27 17:36:30 MY-R: wondering if its a mdev vs udev issue - the initramfs uses mdev during early boot regardless of whether you're using mdev or udev when the full OS is running 2021-10-27 17:38:22 I dont like mdev 2021-10-27 17:38:28 the 1st difference between initial boot and when you replug the adaptor is "usb 2-1: reset SuperSpeed Gen 1 USB device number 4 using xhci_hcd" which appears to be something (udev?) resetting the device 2021-10-27 17:39:35 ye it looks like that 2021-10-27 17:39:50 I have udev on that host 2021-10-27 17:39:59 MY-R: well mdev is what Alpine uses by default in general and don't think anyone has tried running udev in the initramfs 2021-10-27 17:40:07 :/ 2021-10-27 17:40:43 it might just by some mdev configuration that needs tweaked, not sure. 2021-10-27 17:41:08 I like simple things but only when doing their basic job good 2021-10-27 17:41:57 with mdev I had a headache even if wanted do simple partitions on any disk 2021-10-27 17:42:53 they randomly were disappearing when used two different tools etc so stopped using Alpine as rescue distro :/ 2021-10-27 17:44:19 it doesn't make sense for this to be built into the RPI kernel anyway, its not a standard part of the RPI boards so it should be a module - you could perhaps raise an Issue on Gitlab for it to be made a loadable module 2021-10-27 17:45:38 don't use mdev-only systems that much to have noticed issues, the majority of the time I'm working on cloud-init based systems which (currently) require eudev 2021-10-27 17:45:41 rpi configs are taken from raspberry pi kernel github and only lightly modificated by Alpine 2021-10-27 17:48:00 MY-R: that doesn't make them "correct" however. I've been building up a list of suggested changes (mainly switching builtins to modules) for the linux-virt kernels recently getting ready to submit an Issue, haven't got around to the RPI kernels yet as indeed the compile time for testing is slow 2021-10-27 17:51:02 :/ 2021-10-27 17:52:49 https://github.com/raspberrypi/linux/issues/ too many weird issues not mention those closed ones 2021-10-27 17:58:53 Well I'd still recommend you raise an Alpine issue to have that driver changed to be a loadable module, I don't see any point in it being compiled in as (a) only a percentage of RPI owners will have that device and so its being build in unnecessarily increases kernel size, and (b) it have firmware to be loaded which cannot happen until either initramfs is load, if firmware placed there, or rootfs mounted, so no benefit to have it compiled in 2021-10-27 18:01:41 minimal: ye will probably do but first will recompile kernel to be sure, so I need to wait for package which will cool down my rpi4 2021-10-27 18:03:03 MY-R: yeah I'm going to use a desktop rather than a laptop (desktop has better cooling) with the aarch64 VM I'll be using for RPI kernel compiles 2021-10-27 18:04:31 minimal: ordered that thing: https://kamami.pl/77615-thickbox_default/aluminiowa-obudowa-do-raspberry-pi-4-model-b-czarna.jpg 2021-10-27 18:05:28 no ventilation hole ye but if wont be enough then can stick some desktop cpu radiator with thermal glue to it and will be more than enough or even fan 2021-10-27 18:10:28 MY-R: how about https://thepihut.com/products/ice-tower-raspberry-pi-4-cpu-cooler 2021-10-27 18:11:54 is cv2 available as a package for python? 2021-10-27 18:12:42 minimal: nah, I hate small fans and need full cover so nobody will push fingers in to it :P 2021-10-27 18:13:21 MY-R: I see on that page they now do a low-profile version and a metal case for it 2021-10-27 18:14:27 can find lot of weird things for rpi4 todays, too bad rpis prices going up 2021-10-27 18:15:19 I prefer my RPIs "bare" and bolted to a mounting "plate" 2021-10-27 18:17:03 minimal: that is my plan too, will make some small tower to put there small switch, psu, stereo amplifier, raspberry pi, external disk etc... 2021-10-27 18:18:56 MY-R: "poor mans blade chassis" here, RPI+POE mounted using standoffs onto an A5-size piece of foamboard ("blade"), each "blade" slide vertically side-by-side into a "chassis" 2021-10-27 18:19:31 sounds and looks cool for sure :) 2021-10-27 18:19:39 SATA disks also mounted on same size foamboard with SATA-USB adaptor going to adjacent RPI 2021-10-27 18:20:51 :P 2021-10-27 18:21:14 Data-PI-Centre 2021-10-27 18:52:15 honestly, having a few pis in my homelab seems like a fun project. 2021-10-27 18:52:23 but I have a few pis and haven't bothered actually doing it 2021-10-27 18:53:46 i need to figure out what to do with my pi, now that there is alpine i really wanna try doing something with it again 2021-10-27 18:53:58 im behind the times yes but its just recently i got chance to buy replacement sd card 2021-10-27 18:56:43 RPIs are useful for always on background tasks like DHCP/DNS/RADIUS/LDAP servers 2021-10-27 18:57:45 yeah im aware i could do that but i can hardly see a point, and i barely moved in this year so i have awful router that i dont really wanna try take control over, so its kinda difficult to do networking now 2021-10-27 19:13:21 if you always have one computer on anyway, just use that 2021-10-27 19:19:09 if you're like me, you don't want the router to do anything except be an AP and forward packets 2021-10-27 19:22:51 tbh im afraid of even going into its interface because it seems like real crap router 2021-10-27 19:23:15 used to have one that would lock itself up in itself and needed factory reset when that hapened 2021-10-27 19:24:12 heh. Moved 2 years ago, and ISP tech said "your cable modem is old. I will get a new wifi router from the truck." 2021-10-27 19:24:24 I said "fine. But a modem. not a router." 2021-10-27 19:24:50 puzzled look... but he did it. ISP -> ISP's modem -> Alpine 2021-10-27 19:26:57 wait modems are still a thing in this day and age? i thought they dissapeared 2021-10-27 19:27:37 DOCSIS modems are still a thing. You just need to ask. 2021-10-27 19:28:23 every mobile phone has a modem in it. yeah, modems are still a thing 2021-10-27 19:28:38 huh thats interesting, i got DSL internet over here so i will need to find router that accepts that 2021-10-27 19:28:40 Alternative (in USA) is to call Customer Service first thing and ask them to turn your router into a modem. 2021-10-27 19:29:14 I think Comcast calls it "pass through" or something 2021-10-27 19:29:17 wifi too for that matter 2021-10-27 19:29:24 +1 2021-10-27 19:30:28 wireless modems (many or most) support the old AT command set, or subset 2021-10-27 19:30:34 bridge mode it's called sometimes as well 2021-10-27 19:31:21 invoked: yeah.... funny how many first-level tech's don't know that fact. :D 2021-10-27 19:31:47 well.. they are probably too young to know there IS an AT command set! 2021-10-27 19:31:49 korner[m]: there are such things as DSL modems... 2021-10-27 19:32:14 yeah ill need to look for stuff with DSL input, it would be nice if we could get it replaced with normal ethernet cable 2021-10-27 19:33:19 korner[m]: are you in Europe or America or Asia? 2021-10-27 19:33:35 europe 2021-10-27 19:33:44 nangel: to be fair, hardware designers have tried their best to hide these interfaces from people 2021-10-27 19:34:18 so i don't blame the kids for not growing up in the bbs era and basically not knowing about any of it 2021-10-27 19:34:28 korner: Vigor 130 or 166 2021-10-27 19:34:34 Anyone use graphical download managers for Linux? And if so, what one do you like best? 2021-10-27 19:35:10 chip designers like qualcomm only want blessed software/firmware to be used 2021-10-27 19:35:38 ill be looking into it in probably long future, im still job hunting, paying off debts, just lot of future but im glad DSL is still relevant so i can get device for it 2021-10-27 19:36:43 invoked: No, I don't blame them. 2021-10-27 19:37:59 korner[m]: only potential issue with DSL modem is some European telcos, like Deutsche Telekom, only supply voice "lines" via SIP and so expect a combined DSL router with SIP ATA (telephone adaptor) to be supplied/used 2021-10-27 19:39:28 minimal: interesting - so can you run kamailio/freeswitch/asterisk to register to DT's servers, or is it locked down? 2021-10-27 19:40:35 We have similar here in US, but I work for a VoIP provider, so have never needed to see the need to test local ISP 2021-10-27 20:52:31 nangel: not sure, when I was in Germany I didn't use DT but I'm assume in theory the likes of asterisk should work - the likely problem would get getting the config settings, the FritzBox routers are the most popular ones in Germany and I think they might have some "magic" autoconfiguration for things like that unlike generic routers 2021-10-27 21:38:05  2021-10-27 21:41:01 \list 2021-10-27 21:43:20 \list 2021-10-28 02:31:44 not sure if right place to ask, but 2021-10-28 02:31:58 if i have a macvlan interface on a guest, i should be able to access that from the host via a router 2021-10-28 02:32:37 for example the host being 192.168.4.253/24 and the gateway that I want to use being 192.168.4.1/24 2021-10-28 02:32:55 and the guest being hosted on 192.168.4.133/24 (which is accessible from other hosts on my network) 2021-10-28 02:33:23 when i do a traceroute on my host, i see 2021-10-28 02:33:25 1 192.168.4.253 (192.168.4.253) 3047.012 ms !H 3048.251 ms !H 3039.866 ms !H 2021-10-28 02:33:25 traceroute to 192.168.4.133 (192.168.4.133), 30 hops max, 46 byte packets 2021-10-28 02:33:39 this would indicate it is attempting to go directly, not via 192.168.4.1 2021-10-28 02:34:07 so i add a route: 2021-10-28 02:34:09 $ sudo ip route add 192.168.4.133 via 192.168.4.1 dev bond0.4 metric 203 2021-10-28 02:34:30 now i do see it going via the router 2021-10-28 02:34:32 1 192.168.4.1 (192.168.4.1) 0.482 ms 0.624 ms 0.464 ms 2021-10-28 02:34:32 traceroute to 192.168.4.133 (192.168.4.133), 30 hops max, 46 byte packets 2021-10-28 02:34:32 2 * * * 2021-10-28 02:34:37 but nothing returns. 2021-10-28 02:36:46 192.168.4.133 is a webserver, that i use to host some things on my intranet basically 2021-10-28 02:37:21 and normally never need to access it from my host, except this once i thought it might be nice to use it for cloud-init, with nocloud to provision user-data 2021-10-28 07:03:32 is it ok to be off topic here? (anyone know of good noise cancelation headphones, and do they need more power to be driven?) 2021-10-28 07:47:04 i use a pair of very tight earbuds to cancel noise ;) 2021-10-28 08:02:25 frag: we have #alpine-offtopic 2021-10-28 08:03:53 ikke cool 2021-10-28 11:37:00 anyone knows what provides command execstack? i cant find it on alpine package lists 2021-10-28 11:46:38 korner[m]: in alpine it seems noting 2021-10-28 11:47:30 weird, btw is there something i can do about musl java trying to run minecraft natively? seems like that game is built on glibc or something so i end up getting error with wrong architecture even thought both are x86_64 2021-10-28 11:47:52 is execstack provided by glibc? 2021-10-28 11:48:38 doesnt seem like it 2021-10-28 11:49:40 prelink 2021-10-28 11:50:15 https://people.redhat.com/jakub/prelink/ 2021-10-28 11:50:41 what is that? 2021-10-28 11:51:12 it provides execstack 2021-10-28 11:51:58 tbh im not sure if its worth for me trying to fix this stuff, im essentially trying to make alpine do native gaming which is kinda iffy on musl systems 2021-10-28 11:53:35 native gaming sounds like getting closed source drivers functioning. 2021-10-28 11:54:21 i have all drivers working 2021-10-28 11:54:29 this is really just musl built java not liking the game 2021-10-28 11:56:00 if its musl related, you could check musl channels on libera 2021-10-28 11:56:58 there isnt much i can do, its trying to load native library that isnt built on top of musl 2021-10-28 14:26:57 Is Alex Post on here? He made a comment on !26895 that I don't understand. 2021-10-28 14:27:01 It has been merged though. 2021-10-28 14:29:05 Sorry, wrong channel 2021-10-28 20:46:55 Hey guys, I want to ask if Alpine Linux supports ram block device (i.e., brd in Linux kernel https://github.com/torvalds/linux/blob/master/drivers/block/brd.c). I used it in Ubuntu by `modprobe brd` but I cannot do that in Alpine Linux with error message "modprobe: module brd not found in modules.dep". Any idea? Thanks. 2021-10-28 20:47:46 Basically, I want to use ram block device as the back device for file systems (e.g., ext4) to improve file system performance. 2021-10-28 20:50:31 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/linux-lts/config-lts.x86_64#L2268 2021-10-28 20:50:35 the module seems to be enabled 2021-10-28 20:50:46 Jeffrey: what does uname -r and ls /lib/modules return? 2021-10-28 20:51:06 Oh 2021-10-28 20:51:14 Jeffrey: It's not even a module, it's built-in 2021-10-28 20:51:25 so you do not need to load the module, it should just work 2021-10-28 20:52:05 uname -r returns `5.10.61-0-virt` and ls /lib/modules returns two directories "5.10.61-0-virt" and "firmware" 2021-10-28 20:52:58 ok, that is good 2021-10-28 20:53:17 blk_dev_ram isn't set for virt though 2021-10-28 20:53:32 oh, good call 2021-10-28 20:54:43 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/main/linux-lts/config-virt.x86_64#L1795 2021-10-28 20:56:12 Thanks so much! Do I need to switch to STANDARD version of Alpine Linux for using ram block device? I used KVM to boot Alpine Linux 2021-10-28 20:57:19 At least to the lts kernel 2021-10-28 20:58:06 It makes sense. Thanks. 2021-10-28 22:18:59 ram device should have identical performance to loop on tmpfs with dio 2021-10-29 03:17:28 hmm this guide mentions https://christine.website/blog/cloud-init-2021-06-04 /dev/zvol/rpool.... 2021-10-29 03:17:33 but that doesn't seem to exist for me on alpinelinux 2021-10-29 03:17:40 i did see this post https://dev.alpinelinux.org/~clandmeter/other/forum.alpinelinux.org/forum/general-discussion/zfs-devzvol-does-not-exist.html 2021-10-29 03:17:55 but that didn't seem to work for me 2021-10-29 03:23:18 i can't verify myself, but the udev rules in that package look correct, so to double check you do have eudev running and trigger it right 2021-10-29 03:24:45 https://carlosneirablog.wordpress.com/2019/09/28/alpine-linux-create-zvols/ well i found another example 2021-10-29 03:24:50 but it doesn't seem to work :( 2021-10-29 03:28:37 i seem to remember having this problem before 2021-10-29 03:28:43 and solving it but i can't remember how 2021-10-29 03:33:34 seem to remember it having something to do with eudev 2021-10-29 03:34:01 make sure you have all the services started 2021-10-29 03:41:01 i did 2021-10-29 03:41:14 hmm, maybe i should try rebooting, should they on a particular runlevel 2021-10-29 04:05:06 hmm even rebooting didn't help 2021-10-29 04:18:33 hmm i honestly don't know what else to tr 2021-10-29 05:02:34 hi 2021-10-29 09:35:18 leah2: trying 'mflag -S 1' (also tried with full path to msg) and even adding '|| echo fail', no errors, seem to go through but does not change flags nor outpuf filename as confimartion... then later it seemed to randomly work with that file but having the same issue with another file im testing .. btw does mrefile do anything other than move the file (so it could be replaced with 'mv')? 2021-10-29 10:07:26 do you have relative file names in seq? 2021-10-29 10:07:57 and mrefile changes the filename 2021-10-29 10:13:39 leah2: full paths in mseq, eg: /1/s/maildir/m/new/1635424200.25554_1.apk (yeah kinda stupid host name) 2021-10-29 10:19:22 maybe make an strace when it doesnt work 2021-10-29 10:19:54 error handling isnt super good, but i cant remember this being problematic 2021-10-29 10:20:38 okok 2021-10-29 13:58:17 is there anyway for packages to have post-install notes? 2021-10-29 13:58:41 I am thinking about gitea (community) as the non-edge version has a database dump issue 2021-10-29 13:58:54 We remind users that a bug was discovered with gitea dump in 1.14.3–1.14.6 and 1.15.0. Database dumps from these versions cause broken fields in the repo_unit and login_source tables causing the issue identified in #16961. Users on 1.14.x must upgrade to 1.14.7 before running gitea dump. If this is not possible and you are affected #17137 provides a new gitea doctor command to 2021-10-29 13:58:56 fix the repo_unit issue: 2021-10-29 13:59:32 yes but it is discouraged 2021-10-29 14:00:11 in this case specifically, why should the message specifically be printed on install? 2021-10-29 14:03:17 not so much install but upgrade 2021-10-29 14:03:43 since the typical process is to take a backup of your database. however taking a backup creates a corrupted version due to the bug 2021-10-29 15:12:27 minimal: hey! thank you for you help yesterday and also good job, you have got right about my issue with USB-Ethernet adapter 2021-10-29 15:13:00 MY-R: so what was the solution? 2021-10-29 15:13:09 minimal: here a link which used: https://io.24hoursmedia.com/tech-notes/help-my-rpi4-thinks-my-usb-ethernet-is-a-storage-device 2021-10-29 15:14:28 minimal: isnt Alpine thing at all, tried with Void linux too and only reseting usb device helping ('usb-modeswitch' app) 2021-10-29 15:14:49 so like a USB 3G/4G adaptor then, I've used usb-modeswitch with them 2021-10-29 15:15:41 ye 2021-10-29 15:15:50 i think that's got something to do with how usb devices get enumerated on rpis 2021-10-29 15:15:59 ACTION could be wrong 2021-10-29 15:16:52 I would be happy to nuke such a things like some storage in usb modems etc :\ 2021-10-29 15:17:04 but sometimes i think there's an engineer at broadcom trolling the public with the way the rpis are designed 2021-10-29 15:17:30 dunno, after reset, that device changing ID numer to +1 ... 2021-10-29 15:17:52 invoked: those sort of multipurpose device they show as a storage device by default and you need usb-modeswitch to see the "correct" device 2021-10-29 15:18:06 ah. 2021-10-29 15:18:15 it is 2357:0600 when boot and after restart is 2357:0601 2021-10-29 15:18:28 so i was basically correct. :) 2021-10-29 15:18:48 probably both are correct :) 2021-10-29 15:20:55 invoked: nothing to do with RPIs, behaves the same on x86/x86_64 2021-10-29 15:21:50 shows you how little i know 2021-10-29 15:22:17 also i don't use usb much. 2021-10-29 15:22:39 ACTION fades into a shrubbery 2021-10-29 15:24:21 good that raspberry still can boot from usb hdd normaly without stucking on that ethernet adapter 2021-10-29 15:26:22 with usb mouses is same, I have got one which was abble to stop boot process cus older PC thought that is harddrive if plug it in different usb port.... ehh 2021-10-29 16:14:41 hi, is it ok to use irssi-1.1.3, I am kinda stuck with it, experimenting to install alpinelinux on old android phone ? 2021-10-29 16:22:47 ok, figured out so far, --> irc.oftc.net port 6697 [Permission denied], from a user account, but works from root account 2021-10-29 16:25:54 ha! irssi-1.1.3 works from root 2021-10-29 16:26:29 need to logout from root user, bye 2021-10-29 18:05:57 hmm!, stomach churn, cannot make even do ping from user account 2021-10-29 18:09:43 any guide/page as howto set network in an alpinelinux user account? 2021-10-29 19:14:06 Hey guys, thanks for the help yesterday. I tried to boot Alpine Linux standard version via KVM/libvirt and want to use Ram block device. The standard version should have config `CONFIG_BLK_DEV_RAM=y` but I cannot find `brd.ko` in `/lib/modules`. Is there another name for brd in Alpine Linux? Thanks. 2021-10-29 19:15:34 Jeffrey: y means it's built-in, not a module 2021-10-29 19:15:39 =m would mean it's a module 2021-10-29 19:16:20 So it's correct that you cannot find it as a .ko file 2021-10-29 19:22:09 Thanks, is there a way I can use this built-in feature? I can find ram0-15 in /dev/. When I use brd as a module, I can set the size for them, but currently I am not sure how to config /dev/ram*. Thanks again. 2021-10-29 19:32:47 Maybe I can use mknod directly? 2021-10-29 19:37:03 Not sure it's possible with brd being built-in 2021-10-29 19:40:55 Seems like it was set to =y for a very long time already 2021-10-29 19:41:42 Jeffrey: just out of curiosity. Isn't it possible to use tmpfs? 2021-10-29 19:42:54 Thanks. As I am using ext4 requires block device as a back, so I think tmpfs cannot work. 2021-10-29 19:43:02 aha, ok 2021-10-29 19:43:43 You could open an issue on gitlab.a.o asking for it set to be a module 2021-10-29 19:55:50 vkrishn: can you be more specific 2021-10-29 19:56:42 Jeffrey: on linux you can mount any file as a disk image using "loop driver" 2021-10-29 19:57:06 it is built into mount command, so you simply run mount disk.img dir 2021-10-29 19:57:21 if using busybox mount may also need -t ext4 2021-10-29 19:57:33 but why do you think you need ext4 anyways 2021-10-29 20:17:23 what I mean is if someone is already using irssi in alpine, not as root but some unpriv user 2021-10-29 20:17:42 i cannot connect to net as an user 2021-10-29 20:17:52 not even ping :( 2021-10-29 20:33:20 that doesn't sound like an irssi issue 2021-10-29 20:36:28 yes 2021-10-29 20:36:34 i figured 2021-10-29 20:37:35 maybe something i missed during setup 2021-10-29 21:14:12 Can't find xinetd or inetd in the repos (using alpinelinux 3.14 on my raspi). What are the alternatives ? 2021-10-29 21:29:07 freeit: what do you want to achieve? 2021-10-29 21:30:00 it's in busybox-extras, btw 2021-10-29 21:32:08 vkrishn: what is the output of id yourusername 2021-10-29 23:41:15 Hello! I'm having an issue with deving on opengl. Whenever I compile my program I get the error: Undefined symbol: glNamedBuffer, I am assuming I need to install a package for that. Would anyone know which one it would be? My compiling flags is -lGL and -lglfw. I have tried with two compilers, gcc and clang both the same error. 2021-10-29 23:50:11 glNamedBufferData is only present with gl 4.5, i think perhaps what you have doesn't support it or something 2021-10-29 23:51:35 I thought that too, but after checking glxinfo it tells me that my opengl version is 4.6 2021-10-30 00:05:34 libGLESv2 has `glBufferData` 2021-10-30 00:05:48 but that's gles, not sure why it's not in the others 2021-10-30 00:06:38 woh right namedbuffer, confused by names again 2021-10-30 00:06:45 i can only find that in glew, slightly renamed 2021-10-30 00:13:01 post your test case and full command line 2021-10-30 00:15:04 glew did fix the error, but should not be a package for that library like there's in arch and other distros? (libglvnd) 2021-10-30 00:15:32 What do you mean by test case? 2021-10-30 00:18:15 Uploaded file: https://uploads.kiwiirc.com/files/05e7fd757ac23099f93348aee666202d/pasted.txt 2021-10-30 00:18:17 The full error ^ 2021-10-30 00:37:45 glew is unrelated to libglvnd 2021-10-30 00:37:51 and the headers are totally unrelated to glxinfo 2021-10-30 00:43:51 Any idea to fix that error? 2021-10-30 00:57:12 pretty sure you need to use glXGetProcAddress to call this function 2021-10-30 00:57:15 not sure if mesa implements it 2021-10-30 01:02:59 it's in the headers provided by mesa-dev in GL/glext.h under #ifdef GL_GLEXT_PROTOTYPES, and it is mentioned in the mesa source, though i don't know enough about it to know what counts as implemented 2021-10-30 01:15:46 including GL/glext.h still gives me the same error sadly, but using glXGetProcAddress solves my issue. Thank you all! 2021-10-30 01:54:48 @ikke: Thank you for the busybox-extra info. I want to create a scan station with scanbd and the configuration instructions (https://riptutorial.com/raspberry-pi/topic/6701/create-a-scan-station-with-scanbd--raspbian-) requests a superdaemon. 2021-10-30 02:09:31 hey so i opened an mr a wk ago - need some more feeback; https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/26434 2021-10-30 02:09:47 is andypost here? 2021-10-30 02:12:37 its for pkg splitting netsurf - gtk/fb, currently pkg only supports gtk 2021-10-30 02:13:53 you can run just make in build, then make install in subpkg functions, instead of building in them 2021-10-30 02:19:17 psykose: i see that makes sense, so build should have make for gtk & fb; and then specific installs in gtk/fb subpkgs 2021-10-30 02:19:32 if it works, yep 2021-10-30 02:20:23 itll be 2 explicit make commands in build (wont be doable with 1 for both gtk/fb) 2021-10-30 02:21:59 yes, i can see that going through the build myself 2021-10-30 02:26:08 thanks for the help, ill try it out 2021-10-30 07:01:44 !info 2021-10-30 07:02:01 !help 2021-10-30 07:14:28 iquit: just ask a question if you need help 2021-10-30 07:16:00 ok 2021-10-30 07:17:03 oh thanks, I am having problem getting sound to work. booted lts from machine and qemu, apk add alsa-utils, speaker-test runs through but no sound heard 2021-10-30 07:18:45 tried using another kernel in which sound is known to work on other distro too, looks like the kernel has supported drivers, but not sure which files / packages are missing 2021-10-30 07:20:23 did you run an update 2021-10-30 07:27:44 yes if you mean "apk update" before "apk add", I am using the latest ditro, 3.14.2 2021-10-30 07:29:18 and setup-alpine too, also tried rc-service alsa start and rc-update add alsa 2021-10-30 07:32:08 also checked that modules are loaded 2021-10-30 07:34:07 did you check with amixer / alsamixer that it's not muted? 2021-10-30 07:34:56 yes, turned master volume to 100 2021-10-30 07:36:31 I have no experience with sound + qemu, so no idea what else could be missing 2021-10-30 07:38:34 oh no, the traveller found us 2021-10-30 07:39:45 thanks anyway 2021-10-30 07:54:22 btw, I also tried qemu with another distro, so qemu sound is known working, observe same result when booted in read machine 2021-10-30 17:07:44 i unplugged/replugged my hdmi (to fix som sound noise), and my display is now brighter, is there a way to reset it or do i need to reboot? 2021-10-30 17:08:17 not using a desktop environment, using dwm wm 2021-10-30 17:12:39 what video driver? 2021-10-30 17:13:09 intel i guess 2021-10-30 17:28:10 smells like tv range 2021-10-30 18:02:47 https://linuxhint.com/display_brightness_commandline/ >> frag 2021-10-30 18:04:00 Another way: ddcontrol 2021-10-30 18:05:41 or install ddcci-dkms 2021-10-30 18:08:51 was hoping for a simple reset :] 2021-10-30 18:08:56 i guess ill reboot soon 2021-10-30 18:09:14 Go the windows way, unplug and plug in again >:-) 2021-10-30 18:10:27 heheh i think that worked ;) 2021-10-30 18:15:09 Wow. 2021-10-30 19:53:49 can anyone cause a "Mail delivery failed: returning message to sender" message to my mail adr, or do they need a passwd? 2021-10-30 19:55:46 That so does not make any sense to me. Rephrase? 2021-10-30 19:56:37 i got a fraud mail, as if i sent it, thus the subject was "Mail delivery failed: returning message to sender" 2021-10-30 19:57:01 i assume its a cheap attemt they managed without my password 2021-10-30 20:25:41 mail senders can be spoofed, and often are 2021-10-30 20:26:27 sigh. why did i just notice this now? 2021-10-30 20:37:44 lopid: does that mean they have my passwd? 2021-10-30 20:37:56 no 2021-10-30 20:39:30 okok :] 2021-10-30 20:39:45 yeah it was very generic mail... 2021-10-30 20:40:18 but i also dont keep old mails online 2021-10-30 20:45:41 you should change your password anyway, there are two many a's in it 2021-10-30 20:47:38 joke? :P 2021-10-30 20:47:48 ;) 2021-10-30 20:48:12 kinda eerie with those mails 2021-10-30 20:48:36 do you not recognise the from address as your own? 2021-10-30 20:50:06 i think the webmail clients said from "Me", but headers says "From: Mail Delivery System" 2021-10-30 20:51:39 yes, it'll be from 'the system', but it'll refer to a mail that was supposedly sent from your real address, hence the bounce 2021-10-30 20:52:13 hence the bounce going to you, that is 2021-10-30 20:52:53 dunno, not too good with email headers 2021-10-30 20:53:19 the mail's usually quoted in the body 2021-10-30 20:54:18 anyway, it'll probably happen again and become annoying, so i'd consider removing that address if practical 2021-10-30 20:55:30 ah i guess it was "from" my adr yeah 2021-10-31 03:18:54 Hey, I recently upgraded Postgres and all of the commands are complaining about a PQmblenBounded symbol that's missing. Is anyone else experiencing this? 2021-10-31 03:19:58 I'm on edge 2021-10-31 03:26:24 commands run with psql? 2021-10-31 03:28:30 Yeah, and other things like initdb 2021-10-31 03:29:51 > Error relocating /usr/bin/psql: PQmblenBounded: symbol not found 2021-10-31 03:30:32 you can check what libpq it links to with ldd, then see if that libpq has the symbol 2021-10-31 03:30:43 it seems to be correct for both for me 2021-10-31 03:33:00 I checked the library and it doesn't seem to have the symbol 2021-10-31 03:33:05 This is strange 2021-10-31 03:34:39 the only out of date libpq is on mips 2021-10-31 03:35:39 it's also possible it didn't update somehow 2021-10-31 03:42:15 Okay so libpq definitely didn't upgrade 2021-10-31 03:42:37 that would do it 2021-10-31 03:42:48 It seems to refuse to update however 2021-10-31 03:43:15 any alpine image for the rpi400? 2021-10-31 03:43:18 It says that a newer version is avaliable, but when I go to upgrade it it does nothing as if it's up to date 2021-10-31 03:43:31 I don't see any errors with -v 2021-10-31 03:45:57 apk fix -vv should show errors 2021-10-31 03:46:12 freq: that's just the normal rpi image 2021-10-31 03:46:20 to my knowledge 2021-10-31 03:46:25 alright 2021-10-31 03:46:39 only reason i ask is because i tried arch and the 400's bootloader was mad 2021-10-31 03:47:23 Okay I see the problem now. So first of all I noticed postgresql-dev wasn't updating. I forced it to update too and it seems that there's an openssl conflict between the two. 2021-10-31 03:48:16 |> "Black Creek" by "Brent Cobb" on "No Place Left to Leave (2006)" 2021-10-31 03:48:22 it sounds like you have openssl-dev in your world 2021-10-31 03:48:40 postgres-dev needs openssl1.1-dev, openssl-dev is 3.0-dev, and they can't coexist 2021-10-31 03:49:07 this is only for -dev mind you, and also why it's generally not a great idea to keep -dev packages around 2021-10-31 03:50:33 Yeah, I think I've learned my lesson lol 2021-10-31 03:50:36 Thanks for the help 2021-10-31 06:42:52 Heyya everyone. Just wanted to say that the recent GNOME is trying to use systemd. 2021-10-31 06:45:23 what do you mean by that 2021-10-31 07:09:00 Well, since i'm using alpine edge, starting gdm resulting black screen. When checking the log, GDM seems trying to use systemd. 2021-10-31 07:10:25 could you post the log 2021-10-31 07:11:25 sure 2021-10-31 07:19:34 here's some cut of Log: https://pastebin.com/raw/HEKeTtTc 2021-10-31 07:22:25 Yonle: apk add mesa-dri-classic 2021-10-31 07:22:31 the error there is missing the i915 so and nothing to do with systemd 2021-10-31 07:22:55 gnome can try to send things to a dbus interface that doesn't exist all it wants, it's not a requirement for that to be there, hence it's a warning 2021-10-31 07:23:05 afaik there are lots of programs that do this same sort of thing 2021-10-31 07:24:00 https://pkgs.alpinelinux.org/contents?file=i915_dri.so&path=&name=&branch=edge 2021-10-31 08:07:20 hm 2021-10-31 11:58:14 Hey guys, long time alpine (as full desktop) user 2021-10-31 11:58:52 there's an application that I'm trying to compile, which requires qsort_r. This is not available in musl libc as it currently ships, however it was added recently: 2021-10-31 11:59:07 http://git.musl-libc.org/cgit/musl/commit/src/stdlib/qsort.c 2021-10-31 11:59:47 LookingGlassOnA: we already have it included in edge 2021-10-31 12:01:17 Barely a month old, now the application I'm attempting to compile proceeds to the point where it fails due to qsort_r, I wish to see if it would complete successfully if qsort_r is available. My system is running Alpine 3.14.2 (run from ram setup) 2021-10-31 12:01:38 ikke: so should be available in next release then? Great news! 2021-10-31 12:02:26 On my run-from-ram setup, I know messing up the repos and mixing would probably result in a broken system, but can I selectively update musl only to test? 2021-10-31 12:05:26 ikke: say for example, I modify repos, then: 2021-10-31 12:05:32 apk update 2021-10-31 12:05:50 apk add musl=1.2.2-r7 2021-10-31 12:06:11 Sane? 2021-10-31 12:06:44 What release are you on now? 2021-10-31 12:06:53 latest stable (3.14.2) 2021-10-31 12:07:50 It could work, but I'd install busybox-static / apk-tools-static to be sure 2021-10-31 12:13:28 ikke: that did it! I simply issued apk add musl=1.2.2-r7 and only musl and musl-dev were upgraded 2021-10-31 12:13:43 and compile finished successfully! Thanks for the tips 2021-10-31 13:32:49 wwan0 is down after apk update and apk upgrade with 2021-10-31 13:32:49 pmOS v21.06.3 which is supposed to be stable https://pastebin.com/7DNdwA0Y 2021-10-31 13:32:54 Running on a PinePhone 2021-10-31 13:35:17 Might be easiest just to reinstall pmOS but hopefully other people are not having similar issues 2021-10-31 13:38:44 Should have been upgraded from a vanilla PinePhone pmOS image with minor configuration enabling sshd 2021-10-31 13:46:09 Greetings! I am currently running Alpine 3.14.2, which has packaged yggdrasil version 0.3.16-r4. The current release of Yggdrasil is 0.4.0, which has made backwards-incompatible breaking changes. Alpine edge has 0.4.0-r1 in its community repository, which does successfully interoperate with all of my other instances of Yggdrasil 040. I'd rather not have to switch to Alpine edge for this version of Yggdrasil. Would it be possible to get 2021-10-31 13:46:09 it backported? Or, would it be possible to somehow safely grab just this edge package and its edge dependencies and install them on 3.14.2? 2021-10-31 14:38:06 seems like yggdrasil only requires go, busybox, and musl, so you can probably do apk add -X ... add yggdrasil 2021-10-31 14:38:26 in general the safest option is to clone aports and build the package yourself 2021-10-31 14:39:32 or wait for 3,15 unless urgent? 2021-10-31 14:55:11 Not urgent. 2021-10-31 14:55:18 Thanks very much for the suggestions! 2021-10-31 23:54:09 you around, ikke? 2021-10-31 23:55:04 tren: zzz 2021-10-31 23:55:26 clandmeter: you'll do 2021-10-31 23:55:35 when did grsec get removed from alpine? 2021-10-31 23:55:50 few years ago 2021-10-31 23:55:53 is there no way for you guys to just pay the ransom to dipshit brad and distribute it? 2021-10-31 23:56:29 hm 2021-10-31 23:56:30 we are not backed by anything financially 2021-10-31 23:56:38 so it must come from the community 2021-10-31 23:58:21 time for me also to go zzz